Killexams.com 920-330 Dumps and actual Questions
100% actual Questions - Exam Pass Guarantee with high Marks - Just Memorize the Answers
920-330 exam Dumps Source : Communication Server 1000 Rls.5.0 IP Networking Design
Test Code : 920-330
Test designation : Communication Server 1000 Rls.5.0 IP Networking Design
Vendor designation : Nortel
: 60 actual Questions
I exigency actual test questions of 920-330 examination.
I was working as an administrator and was preparing for the 920-330 exam as well. Referring to particular books was making my preparation difficult for me. But after I referred to killexams.com, I create out that I was easily memorizing the pertinent answers of the questions. killexams.com made me confident and helped me in attempting 60 questions in 80 minutes easily. I passed this exam successfully. I only recommend killexams.com to my friends and colleagues for easy preparation. Thanks killexams.
real Q & A of 920-330 exam are top notch!
Killexams.Com provides dependable IT examination stuff, Ive been the disburse of them for years. This exam is no exception: I passed 920-330 the usage of killexams.Com questions/answers and examination simulator. Everything human beings screech is authentic: the questions are genuine, this is a totally trustworthy braindump, completely legitimate. And I actually beget only heard properly things approximately their customer support, however in my view I by no means had issues that might lead me to contact them in the first region. Just first-rate.
These 920-330 questions and answers works in the actual test.
ive searched faultless cloth for this unique matter matter over on line. but I could not find an confiscate one that perfectly explains best the wanted and captious things. whilst i discovered killexams.com understanding dump material i was genuinely amazed. It simply included the essential things and not anything crushed within the dumps. i am so excited to locate it and used it for my coaching.
Questions had been precisely identical as i purchased!
The inquiries are valid. basically indistinguishable to the 920-330 exam which I passed in just 30 minutes of the time. If notindistinguishable, a extremely benign deal of stuff could live very an terrible lot alike, so that you can conquer it supplied for you had invested enough making plans energy. i used to live a bit cautious; however killexams.com and examSimulator has became out to live a tenacious hotspot for examination readiness illumination. Profoundly proposed. thank youso much.
Did you attempted this notable source cutting-edge dumps.
Many thanks for your 920-330 dumps. I recognized maximum of the questions and besides you had plenary the simulations that I was asked. I were given 97 percentage score. After trying numerous books, I was pretty disenchanted no longer getting the proper substances. I become searching out a guideline for exam 920-330 with easy language and properly-prepared content material. Killexams.Com fulfilled my want, because it defined the complicated topics within the simplest way. In the actual exam I were given ninety seven%, which changed into beyond my expectation. Thank you killexams.com, to your tremendous manual-line!
Very easy to accept licensed in 920-330 examination with this examine guide.
Within the wake of attempting some aids, I at closing halted at Dumps and it contained precise answers added in a easy manner that was exactly what I required. I used to live struggling with with topics, while my examination 920-330 become handiest 10 day away. I used to live paralyzed that i might now not beget the capacity to gather passing rating the lowest pass imprints. I at remaining handed with 78% marks without an terrible lot inconvenience.
Extract simultaneous plenary 920-330 path contents in layout.
I got 79% in 920-330 Exam. Your study material was very helpful. A tremendous thank you kilexams!
920-330 bank is required to limpid the exam at the nascence try.
Remarkable 920-330 stuff, 920-330 telling questions, 920-330 rectify answers. Expert exam simulator. I used to live relieved to observethat this schooling percent. Has essential facts, genuinely what I had to realize to skip this exam. I abominate whilst they are trying to promote you things you dont want in the first vicinity. This wasnt the case despite the fact that, I had been given exactly what I needed, and that is validated by using the fact that I passed this 920-330 exam ultimate week, with a nearly best score. With this exam experience, killexams.Com has obtained my deem for years yet to come.
Nice to hear that dumps of 920-330 exam are available.
To become a 920-330 certified, i was in push to skip the 920-330 examination. I tried and failed closing 2 tries. Accidently, I had been given the killexams.Com fabric through my cousin. I was very inspired with the fabric. I secured 89%. I am so joyful that I scored above the margin note with out problem. The fabric is well formatted as well as enriched with crucial principles. I reflect it is the extremely benign wish for the exam.
it's miles remarkable to beget 920-330 actual exam questions.
i was looking to accept ready for my 920-330 check that became across the nook, i discovered myself to live lost within the books and wandering far far from the actual point. I didnt understand a solitary phrase and that turned into virtually concerningdue to the fact I had to prepare as soon as feasible. Giving up on my books I decided to register myself in thisKillexams.com and that changed into the pleasant decision. I cruised through my 920-330 select a spy at and became capable of accept a respectable score so thanks very an terrible lot.
Whilst it is very hard stint to pick trustworthy exam questions / answers resources regarding review, reputation and validity because people accept ripoff due to choosing incorrect service. Killexams. com execute it certain to provide its clients far better to their resources with respect to exam dumps update and validity. Most of other peoples ripoff report complaint clients approach to us for the brain dumps and pass their exams enjoyably and easily. They never compromise on their review, reputation and attribute because killexams review, killexams reputation and killexams client self assurance is considerable to plenary of us. Specially they manage killexams.com review, killexams.com reputation, killexams.com ripoff report complaint, killexams.com trust, killexams.com validity, killexams.com report and killexams.com scam. If perhaps you perceive any bogus report posted by their competitor with the designation killexams ripoff report complaint internet, killexams.com ripoff report, killexams.com scam, killexams.com complaint or something like this, just maintain in understanding that there are always horrible people damaging reputation of benign services due to their benefits. There are a great number of satisfied customers that pass their exams using killexams.com brain dumps, killexams PDF questions, killexams practice questions, killexams exam simulator. Visit Killexams.com, their test questions and sample brain dumps, their exam simulator and you will definitely know that killexams.com is the best brain dumps site.
Back to Braindumps Menu
000-287 mock exam | 00M-624 actual questions | C2010-518 free pdf download | C2090-320 practice questions | HP0-058 exam questions | 3M0-331 dumps questions | HP0-M51 test prep | 920-338 free pdf | 300-360 study guide | 000-298 exam prep | HP2-B71 study guide | HP0-055 questions answers | OMG-OCUP-100 brain dumps | P2020-007 bootcamp | HP2-K40 dumps | 000-042 dump | 000-R25 actual questions | ISSAP actual questions | C2180-317 test prep | 000-417 sample test |
920-330 exam questions | 920-330 free pdf | 920-330 pdf download | 920-330 test questions | 920-330 real questions | 920-330 practice questions
killexams.com 920-330 Communication Server 1000 Rls.5.0 IP Networking Design exam brain dumps with practice programming.
We are for the most participate very much awake that a noteworthy issue in the IT industry is that there is an absence of value study materials. Their exam prep material gives you plenary that you should select a certification exam. Their Nortel 920-330 Exam will give you exam questions with confirmed answers that reflect the actual exam. high caliber and incentive for the 920-330 Exam. They at killexams.com are resolved to enable you to pass your 920-330 exam with high scores.
We beget Tested and Approved 920-330 Exams. killexams.com provides the most accurate and latest IT exam materials which almost accommodate plenary lore points. With the aid of their 920-330 study materials, you dont exigency to blow your time on reading bulk of reference books and just exigency to disburse 10-20 hours to master their 920-330 actual questions and answers. And they provide you with PDF Version & Software Version exam questions and answers. For Software Version materials, Its offered to give the candidates simulate the Nortel 920-330 exam in a actual environment.
killexams.com Huge Discount Coupons and Promo Codes are as under;
WC2017 : 60% Discount Coupon for plenary exams on website
PROF17 : 10% Discount Coupon for Orders greater than $69
DEAL17 : 15% Discount Coupon for Orders greater than $99
OCTSPECIAL : 10% Special Discount Coupon for plenary Orders
If you're searching out Pass4sure 920-330 practice Test containing actual Test Questions, you are at privilege vicinity. They beget compiled database of questions from Actual Exams so as that will assist you effect together and pass your exam on the first attempt. plenary schooling materials at the website are Up To Date and proven with the aid of their specialists.
We tender ultra-modern and up to date Pass4sure practice Test with Actual Exam Questions and Answers for brand modern syllabus of Nortel 920-330 Exam. practice their actual Questions and Answers to help your expertise and pass your exam with high Marks. They execute certain your pass inside the Test Center, protecting plenary of the subjects of exam and construct your lore of the 920-330 exam. Pass four certain with their accurate questions.
killexams.com 920-330 Exam PDF includes Complete Pool of Questions and Answers and Dumps checked and confirmed inclusive of references and causes (where relevant). Their target to collect the Questions and Answers isn't always best to pass the exam at the start strive but Really help Your lore about the 920-330 exam topics.
920-330 exam Questions and Answers are Printable in high attribute Study usher that you may down load in your Computer or another device and start preparing your 920-330 exam. Print Complete 920-330 Study Guide, deliver with you while you are at Vacations or Traveling and delight in your Exam Prep. You can accept privilege of entry to up to date 920-330 Exam out of your on line account anytime.
killexams.com Huge Discount Coupons and Promo Codes are as under;
WC2017 : 60% Discount Coupon for plenary assessments on website
PROF17 : 10% Discount Coupon for Orders extra than $69
DEAL17 : 15% Discount Coupon for Orders more than $ninety nine
OCTSPECIAL : 10% Special Discount Coupon for plenary Orders
Download your Communication Server 1000 Rls.5.0 IP Networking Design Study usher without delay after shopping for and Start Preparing Your Exam Prep privilege Now!
920-330 Practice Test | 920-330 examcollection | 920-330 VCE | 920-330 study guide | 920-330 practice exam | 920-330 cram
Killexams 70-567-CSharp free pdf | Killexams RHIA VCE | Killexams HP0-J63 braindumps | Killexams 250-223 test questions | Killexams 000-N15 test prep | Killexams 250-253 exam prep | Killexams 500-452 brain dumps | Killexams HP0-J18 pdf download | Killexams VCS-272 dump | Killexams 6005-1 study guide | Killexams 190-821 practice test | Killexams BMAT braindumps | Killexams 920-245 mock exam | Killexams 156-215.13 free pdf | Killexams BCP-620 questions and answers | Killexams 000-599 sample test | Killexams 9L0-008 brain dumps | Killexams FINRA exam questions | Killexams VCS-371 practice questions | Killexams 050-634 study guide |
killexams.com huge List of Exam Study Guides
Killexams F50-532 exam questions | Killexams ISEB-SWT2 questions and answers | Killexams 9A0-148 VCE | Killexams 920-551 free pdf | Killexams 000-563 study guide | Killexams 050-663 exam prep | Killexams 250-412 actual questions | Killexams BMAT bootcamp | Killexams HP2-H01 dump | Killexams 000-370 actual questions | Killexams JN0-541 practice questions | Killexams 650-179 free pdf | Killexams 9A0-055 questions and answers | Killexams 000-101 study guide | Killexams M2020-229 cram | Killexams HP0-S44 cheat sheets | Killexams 350-023 braindumps | Killexams HP3-042 questions answers | Killexams 1Z0-520 pdf download | Killexams FN0-202 mock exam |
Communication Server 1000 Rls.5.0 IP Networking Design
Pass 4 certain 920-330 dumps | Killexams.com 920-330 actual questions | https://www.textbookw.com/
Asynchronous communication is a widely-used communication method between different processes and systems. In an asynchronous communication, the client sends a request to the server (which requires lengthy processing) and receives a delivery acknowledgment privilege away. Different from the synchronous communication, this response does not beget the required information, yet.
After the client receives the acknowledgment, it continues to accomplish its other tasks, assuming it will eventually live notified when the required information is ready on the server side.
The biggest benefit of asynchronous communication is the increased performance. Since the client does not secrete its valuable CPU cycles just for waiting, it can deliver more within the identical timeframe. Increased decoupling between the client-server interaction will besides lead to better scalability.
We perceive asynchronous communication patterns everywhere. Here are some examples:
A “Design and Assign” request is submitted to the Inventory Management Application, from the Order Management Application.
A “full dump” is requested from an Inventory Management Application.
Monitoring Application sends 1000 SMS’s to the service impacted customers via an SMS Gateway.
Examples can live multiplied, but the principle is the same: Notify the caller when the lengthy process is finished, and the information can live consumed.
There are three methods to implement an asynchronous communication:
Using Pub-Sub messaging using a Message Broker (or MoM)
Polling for situation Changes
In this article, they will intricate these methods and some strategies that can execute them effective.
Method 1: Asynchronous Callback
In an asynchronous callback mechanism, following steps occur:
The client authenticates to the server.
The client calls the server operation. (Web service, RPC, Local method call, etc.)
The client besides subscribes with its “callback endpoint address” to the server. (explained below)
The server acknowledges the receipt of the request synchronously.
Client waits for the reply from another pre-defined channel (A Servlet, PHP page, Local handle, etc.)
Server finishes the required drudgery and notifies the client from the channel.
The client fetches the information and processes it.
Method 2: Broker-based Publish/Subscribe
In this method, a “topic” is created to enable the Client-Server communication. The steps are similar to Asynchronous Callback, but here, the medium differs. The server never notifies the client directly. It does this through a buffer, which is the Broker.
The client authenticates to the server.
The client calls the server operation. (Web service, RPC, Local method call, etc.)
The client subscribes to the broker and starts listening to the topic from a different thread.
Server finishes the required drudgery and publishes a message to the topic.
The client fetches the information and processes it.
Since they faith on a different broker component that will accomplish the mediation between the systems, they should beget a solid understanding of the inner workings of that broker. Features like message durability, TTLs, and routings exigency to live elaborated thoroughly.
Method 3: Polling
Polling should live the least preferred method from the performance and scalability point of view as it puts extra strain on both client and server side. However, in some conditions, (especially when you beget no control over the legacy server application’s code or repository), you may live forced to implement it. Here are the typical steps of polling:
The client authenticates to the server.
The client calls the server operation. (Web service, RPC, Local method call, etc.)
Server acknowledges the receipt of the request synchronously. Server puts the request in its database or exposes its situation via an external service (such as web service)
Every X seconds, client polls the situation of the request by connecting to the repository or the exposed interface.
If request’s situation transitions to “ready”, the client fetches the information and processes it.
There are certain strategies you exigency to deem while designing asynchronous communication architectures.
The participants should live able to uniquely identify each request. That is to say, if the client asks the server to dump its database to an FTP server, the server should revert its acknowledgment with a key that identifies this individual request.
The client can, then, wait for this particular key in its listening channel and correlate the incoming notification to the original request. Ideally, this key should live generated by the server. However, in some situations (cloud trailing requirement or legacy application involvement), the client provides a unique key attached to the request. It is then the server’s responsibility to respond back with the identical key when the callback time comes. The drawback to this second approach is key collisions. If a part client besides provides the identical key at the identical time, the server will exigency to reject the request.
Broker-based Publish/Subscribe method normally uses one shared topic for plenary clients. Key Strategy becomes extremely considerable especially when this method is chosen.
Imagine you are implementing the callback approach with an external URL. The remote client has passed the request, got its acknowledgment and waiting for the callback event to live delivered. What if the clients’ endpoint is not available at that instant due to some reason? (Network outage, rebooting due to patch deployment, etc.)
If the server simply ignores this callback, when the client comes back up, it will never receive the callback. Therefore the request will never live fulfilled; client resources would live unnecessarily consumed.
To avoid this situation, Server should implement retries. It should retry the callback multiple times, waiting for fixed/increasing intervals in between. If the remote participate never comes alive, then the callback message can live effect in a repository that can live “re-played” manually by the support personnel.
With the broker approach, retry strategy can live even more challenging. There is a dusky side of the publish/subscribe model. When you publish a message, it will live delivered to plenary of the subscribers. If the subscriber is not listening at that instant though, the message is lost! There are some workarounds to avoid this situation, such as durable application server topics, attached queues, or some tools like Apache Kafka. please note that these workarounds can approach with increased costs of maintainability, so feasibility studies should live performed before the rollout.
Asynchronous Callback method requires a subscription strategy. The client should provide the server its address. For webhooks, this is a URL hosted on the client’s web server. For other cases, it could even live a hostname and port number.
Rather than putting client URLs to a central database before the integration starts, they should implement a dynamic endpoint subscription methodology. The modern course to accomplish this is to provide a Restful web service endpoint which accepts a request id, URL, and a key. “request id,” comes from the initial synchronous request they made, which will live used as the correlation key. “URL” is the client’s callback address. “key” is the password that should live passed to the client along with the URL callback.
Before the callback happens, the server can spy up a “request id” from a lookup table (fed previously by a subscription) and find out the endpoint address to call. If this is a one-off request/response pair, the lookup row can live deleted from the repository on the spot.
Generated response on the server side can portray any information. It can live a ten digit number or a ten terabyte file. Payload strategies depict how this information is passed to the client side.The payload can live directly passed inside the asynchronous notification itself. If the size is expressed in kilobytes, they can pass the information along to the callback. If this is not the case, the pointer to the file should live passed in the notification. If the information is captured in a, say, ten terabyte file, a file name, and an FTP server IP address can live passed within the notification. It would live then the client’s responsibility to Go ahead and fetch that file.
Designing asynchronous systems require careful design. The first question they exigency to query ourselves is “Will it live more feasible to accomplish this synchronously?”. If the non-functional requirements allow, they should stick to the synchronous course of doing things. If you End up deciding the asynchronous path though, methodologies and strategies that are mentioned in this article can execute your journey smoother.
performance ,asynchronous communication ,asynchronous ,asynchronous programming ,asynchronous methods
WESTFORD, Mass.--(BUSINESS WIRE)--Sonus Networks, Inc. (Nasdaq: SONS), a global leader in SIP communications, today introduced modern software and hardware enhancements to the Sonus SBC 1000 and SBC 2000 Session rim Controllers (SBCs) which provide customers greater investment protection by enabling increased session and port capacity in a quick, simple manner. With the continued rapid growth of modern real-time communications applications such as Microsoft Lync Enterprise Voice, there is a corresponding claim for more session throughput in customer deployments. Today’s announcement simplifies how enterprises – and the resellers who support them – can scale session capacity to meet the needs of Voice over Internet Protocol (VoIP), video and other latency-sensitive collaboration tools associated with Unified Communications (UC) deployments.
With the introduction of Release 3.2, increased session and port capacity on the SBC 1000 and SBC 2000 can now live activated remotely with a license key, removing the exigency for technician-assisted installation of additional hardware. The supple architecture of the SBC 1000 and SBC 2000 allows service providers, resellers and enterprises to quickly add capacity without the inconvenience of service disruption. Sonus is the only brand to tender such ease-of-use capability in a arm office SBC portfolio.
As workplace communications increasingly select space outside of formal office locations, Session Initiation Protocol (SIP)-based communications beget become foundational as companies strive to deliver the identical capabilities regardless of where an employee works. SBCs serve a captious role in securing and enabling disparate technologies across service provider and enterprise networks to drudgery together seamlessly so that Microsoft Lync and similar collaboration platforms can deliver real-time communications such as VoIP, video, presence, unified messaging, find me/follow me, virtual whiteboarding and document sharing. By integrating Sonus SBCs into real-time communications deployments, these powerful network devices protect, secure, simplify and standardize what otherwise would live very complicated SIP-based multimedia communications.
By helping customers future-proof their networks without over- or under-subscribing to session or port capacity at the time of deployment, the modern configurations of the SBC 1000 and SBC 2000 address today’s dynamic industry environments. Customers can purchase the capacity they exigency today with assurance that increased session and port matter can live added in minutes, instead of hours or days, through the disburse of a software license key. The expanded control over network growth empowers service providers and enterprises to simplify network design, reduce costs and more effectively deploy advanced real-time communications features.
The Sonus SBC 1000 and SBC 2000 are integrated into the Sonus ingredient Management System (EMS), delivering one common network management solution across the complete Sonus SBC product portfolio. This release amplifies EMS functionality on the SBC 1000 and SBC 2000 with support for network-wide provisioning, scheduled backup/restore and scheduled updates to assist customers reduce operating costs, optimize investments and maximize network performance.
Quotes:“Customers pick The Via Group based on their skill to assist them deploy and leverage technology in a course that makes their industry more productive. It is raining Lync upon us privilege now, so anything they can accomplish that quickly scales their back-end operational efficiency and enables The Via Group to disburse more time focused on adding value to a customer’s undergo versus simply managing technology is highly beneficial,” said Michael Cassady, director of operations, The Via Group. “The skill to remotely expand session and port capacity with the Sonus SBC 1000 and SBC 2000 makes these products a foundational ingredient to real-time communications and large-scale, multi-site customer deployments.”
“The days of reconfiguring SBCs with a truck-roll to accommodate growth of session or port capacity are a headache of the past. With the SBC 1000 and SBC 2000, increasing session or port capacity is as simple as downloading a license key,” said David Tipping, vice president and general manager, SBC Business, Sonus. “With Release 3.2 of the SBC 1000 and SBC 2000, service providers and resellers can meet customer claim more rapidly, remove unneeded operational friction, and ensure investment protection on behalf of their customers.”
As a Microsoft Gold Communications Partner, the Sonus SBC portfolio spans from the arm to the network core. It is the only brand offering enterprises and service providers a complete end-to-end, Lync-qualified solution.
For consecutive years, Sonus has been positioned in the Leaders quadrant of the “Magic Quadrant for Session rim Controllers,” published October 21, 2013 by Gartner, Inc.
Sonus SBC solutions are Miercom Performance Verified and are deployed in many of the world’s mission-critical networks spanning finance, oil and gas, manufacturing, transportation, education, retail, healthcare and pharmaceutical markets.
Click here to download electronic copies of Sonus’ industry-leading reference guides: Lync Enterprise Voice for Dummies, SIP Trunking for Dummies, Session rim Controllers for Dummies and WebRTC for Dummies.
Click here to learn how Sonus realized productivity gains and significant revert on investment (ROI) from its own Lync Enterprise Voice deployment.
View complete information about the Sonus SBC portfolio at www.sonus.net.
Tags/Keywords:Sonus Networks, SONS, Lync, real-time communications, Session rim Controllers, SBC, UC
About Sonus Networks:Sonus helps the world's leading communications service providers and enterprises embrace the next generation of SIP-based solutions including VoIP, video and Unified Communications through secure, trustworthy and scalable IP networks. With customers around the globe and 15 years of undergo transforming networks to IP, Sonus has enabled service providers and enterprises to capture and retain users and generate significant ROI. Sonus products comprise session rim controllers, policy/routing servers, subscriber feature servers and media and signaling gateways. Sonus products are supported by a global services team with undergo in design, deployment and maintenance of some of the world's largest and most complicated IP networks. For more information, visit www.sonus.net or muster 1-855-GO-SONUS.
Important Information Regarding Forward-Looking Statements:The information in this release may accommodate certain forward-looking statements within the sense of the U.S. Private Securities Litigation Reform Act of 1995 regarding future events that involve risks and uncertainties. Although Sonus believes that its expectations are based on reasonable assumptions, readers are cautioned that these forward-looking statements are only predictions and are matter to inherent uncertainties, risks and changes in circumstances that are difficult to predict. plenary statements other than statements of historical facts contained in this report are forward-looking statements. Their actual results may vary materially from those contemplated by the forward-looking statements. For further information regarding risks and uncertainties associated with Sonus' business, please advert to the "Management's Discussion and Analysis of monetary Condition and Results of Operations" and "Risk Factors" sections of Sonus' filings with the Securities and Exchange Commission. Any forward-looking statements portray Sonus' views only as of the date on which such statement is made, and should not live relied upon as representing Sonus' views as of any subsequent date. While Sonus may elect to update forward-looking statements at some point, Sonus specifically disclaims any duty to accomplish so, except as required by law.
Marvell Technology Group Ltd. (NASDAQ:MRVL) Investor Day Conference October 16, 2018 9:00 AM ET
Ashish Saran - VP, IR
Matt Murphy - President and CEO
Dan Christman - EVP of Storage Group
Raghib Hussain - EVP and Chief Strategy Officer
Tom Lagatta - EVP of Worldwide Sales and Marketing
Jean Hu - CFO
Ross Seymore - Deutsche Bank
Blayne Curtis - Barclays Capital
Karl Ackerman - Cowen & Company
John Pitzer - Credit Suisse
Vivek Arya - Bank of America Merrill Lynch
Quinn Bolton - Needham & Company
Good morning, folks, and welcome to Marvell's 2018 Investor Day. For those who don't know me, my designation is Ashish Saran; I'm the Vice President of Investor Relations at Marvell. I've been in the semi industry for over 20 years, which should execute you question my sanity, but putting that aside, as some of you know, I recently joined Marvell. I was attracted by the very significant growth opportunities I perceive in front of this company, especially with the addition of Cavium. I'm besides very pleased to perceive a lot of close faces in the audience, so thank you everyone for taking time out of your diligent schedules and spending today morning with the Marvell team.
As you can see, they beget a very informative day in front of us, so Matt is going to kick things off with an update on their strategic shift to infrastructure. Dan and Raghib are going to walk you through the nuts and bolts of their storage and networking businesses. Tom will entertain you with an update on their go-to-market strategy to drive growth. And as anyone who knows Tom can attest, "Entertain" is the privilege word to disburse in describing Tom's style. Jean will deliver the money slides, followed by a mp;A session which will End the event.
Now, before they start, I accomplish exigency to select you through their very exciting monetary safety briefing. This presentation today will accommodate certain forward-looking statements which accomplish beget risks and uncertainties. They recount these in their filings with the SEC. They will besides live mentioning certain non-GAAP monetary measures, a reconciliation is at the End of this presentation, and this presentation will live available on their Web site after today's event.
Now that plenary of you beget your seatbelts tightly fastened, we're going to kick things off with a short video, followed by Matt's presentation. Thank you.
The world today, it's more demanding than ever because it's more connected than ever. Manufacturers are more autonomous because they're more automated. Smart cities provide vital services thanks to tracking data and voice technology. A solitary car can learn how to drive safely, then participate it with 10 million others. And soon, highly intellectual 5G base stations will connect to billions of users at breakneck speed. These emerging applications are powered by learning machines, where information is sent to the core for analysis, then back to the edge, flooding the global network with more traffic than rush hour in the rain.
To maintain it flowing, the datacenter needs to maintain morphing, so much so that you can't expose where the core ends and the edge begins. The solution rests in the foundational technology here today. They understand; they helped build it. With decades of experience, the combined portfolios of Marvell and Cavium beget broadened their capabilities, made their core strengths even stronger, and transformed us into an infrastructure powerhouse. We've created seamless bandwidth connections between the core datacenter and the network edge, proper entire computing systems on to a solitary microchip, increased storage density so companies can meet tomorrow's demands, not live surprised by them, and discovered ways to maintain data flowing swiftly and securely, through cables or through the air. The modern Marvell has its sights on tomorrow, not just to unearth what's next, but to execute it available today.
Marvell, they reflect ahead, so their customers can too.
Okay. benign morning, everybody. It's considerable to perceive plenary of you here. So I think, first of all, this video says a lot about what's going on today. The situation of technology and it's mind-blowing how much data is being created on a daily basis in the world. If you reflect about it, most of this data today is being moved into the cloud, but more and more the data is actually being created at the edge where it needs to live processed and actioned. And Marvell is enabling the infrastructure that makes plenary of this possible. So today I'm going to talk about the kind of company that we're creating, how far we've come, where we're going, and how we're doing since we've combined with Cavium. And I'm going to expose you how we're going to grow, which I'm certain is on everybody's minds here today.
So first, let me expose you how far we've approach from their last investor day, which was just six quarters ago, here in this room. So let me select you back, this was in March, 2017. For those of you that were here at the time, it was snowing outside. And this was their first ever investor day for the company. I had been on the job for about eight months, and they were just nascence Marvell's turnaround. We'd assembled a modern leadership team which they introduced everybody, and they told their story. And that tale started with the market opportunity, which was really fueled by the explosion of data and the exigency for bandwidth. They said they were shifting from consumer and mobile to cloud and infrastructure, with the depth of their IP and core capabilities really played to their strengths, and where they felt they could create the most value.
We besides shared their plans at that time how they were going to refocus the company to build a long-term sustainable industry that could deliver vigorous margins, cash flow; plenary the things that you await from a top-tier semiconductor company. And I reflect at that time they caught some of you off guard with their transparency. I recall during the mp;A, I reflect it was Chris Roland, who I reflect is in the room here, who said, "I'm silent trying to pattern out what company this is. I perceive the logo behind you, it says Marvell, so I guess I will just Go with that." So they did set some very high expectation for ourselves in that meeting. And I'm joyful to report that the team has done a considerable job on delivering on their commitments.
So let's start with revenue growth, the chart -- every chart here starts with Q1 '18, which was the quarter after they had their analyst day. You can perceive that revenue since that time has been up into the right. They besides said at that time that they would achieve grievous margins exceeding 60% exiting their fiscal '20. Really pleased that subsequent to the analyst day they managed to enlarge their grievous margins every quarter. In the last reported quarter they had in Q2, they reached 63.5%, which was an all-time record for Marvell. And finally, they had committed to deliver 30% operating margins existing fiscal '20. And again in Q2, the last one they reported as a standalone company, they besides exceeded 30% OM, which was about six quarters ahead of schedule.
So these were considerable results. We're very proud of them. But they did not betide by accident. So let me talk now about how they elude their business. So we've really established a very results-oriented culture inside the company. It's really based on data-driven decision-making and it reflects my core beliefs and values as a leader. They start by first aligning to the privilege markets. They are focused on fewer things, but we're focused on doing those fewer things very well. They actually hold a portfolio review of plenary of their businesses in the combined company. We've done three of these now since I've joined. The last one, more recently, was with Cavium. And so we've gone through now every industry in the combined company.
We understand where every R&D dollar is being spent; they understand the competitive dynamics of those businesses, the IRRs, the grievous margins, the growth potential, the competitive landscape. And we've now built P&Ls and three-year plans for each of those businesses. So we've actually integrated Cavium in quite quickly into their planning process. We're besides not fearful to halt projects or redirect resources or reallocate their precious R&D dollars where they beget to. We're very disciplined in this front. They besides believe that time-to-market is a differentiator in their industry. So, their customers, they draw their schedules around their equipment around their schedules. They're plenary one and the same. And so we've completely retooled inside the company how they draw their projects, their engineering projects, how they execute them, how they track them, and how they bring them to production.
So we've installed world-class program management, and most importantly, a culture of accountability inside the company. And their customers are seeing the difference. They besides believe in a very data-driven approach to industry management. And they disburse data to elude plenary aspects of their business, whether that's pricing, forecasting, claim planning, or measuring customer satisfaction. And we're making better decisions because of it, and we're already seeing the results. So this has been a very programmatic, thoughtful effort. And they perceive the benefit of this character of approach on the modern Cavium team coming in.
Part of the turnaround besides in the company though has been cultural. And over the past two years we've really brought the entire company along with us. Through frequent and transparent communications we've really strengthened the culture inside the company. And it starts with articulating their core behaviors and what they value and stand for. And these behaviors, by the way, they're not just coffee mug slogans and posters they stick on the wall; these are very personal to me. And it really reflects how they elude the company. Integrity, respect, innovation, execution, supporting each other, these are plenary very powerful words that resonate very well with their employees and their customers, by the way.
And we've now received numerous supplier awards in the last two years. And recently they made the Forbes Best Employers list in 2018. So they believe culture is a competitive advantage. The employees perceive the tremendous picture, they understand their role, they'll buy in and they'll fully entrust when they buy into your culture. And so this photo, by the way, is a photo of the combined Marvell and Cavium management team. It was taken shortly before they closed the merger. They plenary took time to accept offsite and accept to know each other and align on their goals. And by the End of that offsite, you couldn't expose who was a Cavium leader and who was a Marvell leader; we've plenary become one team. And I would note that in this picture about 30% of the people, or almost one-third of the vice presidents in the company combined, are actually from Cavium.
So they truly merged the companies, versus just acquiring and taking the products. And I couldn't live more proud to plod forward with such a talented team that we've assembled. And they set their mission to live the leading semiconductor company serving the infrastructure market. One that developed solutions that move, store, process, and secure the world's data faster and more reliably than anyone else. They believe it's considerable to beget a mission that is easy to understand. Everyone can accept behind it whether it's an employee, whether it's one of their customers, or everybody in the room.
So let's select a instant to talk about why they like infrastructure; why is this a benign market to live in? I reflect it's an attractive long-term market where their IP can deliver actual value, and so let me contrast market. So there's the consumer market, is exciting. great TAM, you can talk about it with your friends at a party, it's easy to understand. But I'm telling you, as quick as success comes in this market as quick as it can Go away. And this is not just an academic observation I'm making, I've lived this. I've been in this world. I've lived through the notebook cycle, the digital camera cycle, the smartphone cycle, the smart TV cycle. This is a brutal industry to live in. And they are fortunate at Marvell to beget the frill to participate in the infrastructure market. This is one where they don't just beget to re-win designs every year to breakeven and wait on the treadmill. The design cycles and infrastructure are typically three to five years. So there is a long cycle of development and partnership with your customers. You don't exigency to start over every year. And those wins that you get, you actually layer on top of each other. So they compound over time.
These platforms are typically in the bailiwick for 10 years or more. So, when customers spy at who they are going to ally with and design in, there's really only a handful select group of players that really beget plenary the capabilities that they are looking for. So infrastructure is a much more predictable stable market. And that benefits everybody. It benefits their customers, their shareholders, and their employees who can focus on the long term as they develop their products. besides the company's DNA is very well aligned to the infrastructure market. They beget a tenacious combination of IP and engineering capabilities in the company. They beget a 20-year track record in Ethernet, both in switches and in PHYs. They beget profound undergo with processor cores. And they beget in-house IP development such as SerDes and others that they control their own destiny on.
Finally with Cavium, they bring in a very compelling software, plenary platform that they can leverage across multiple product lines and really add a lot more value and stickiness to their customers. And finally, they beget managed to assemble on this combine company really what I can screech is the world's most talented team of blend signal engineers and digital engineers designing SOCs. These chips are not easy to make. There's only a few companies that possess this character of capability. And we've been a top innovator in their field. They beget been named for the last six years as one of the top innovators in the world by Thomson Reuters and now Clarivate. And today, the company has amassed over 10,000 patents, which is very formidable and is just another instance of the kind of capability that they have. So on their last Investor Day they said they were pivoting the infrastructure and that it would live a journey. A few years back, they were about 25% of the company's revenue with an infrastructure of 75% was in things like consumer and mobile.
At the last Analyst Day, they were at about one-third of their revenue was levered infrastructure. And they said their goal was to accept it to 50% in the near-term. And so, they were able to accomplish that actually a cramped bit ahead of schedule through plenary the drudgery that they effect in. And so, classic Marvell as of today is about half-half, fifty percent infrastructure, 50% non-infrastructure. So that was progress. They were there probably a year or so ahead of where they thought. Cavium really represented the next step in their progress here. And so when you combine Cavium in, the combine company today is now about two-thirds of their revenue is levered to these types of End markets. And that's only going to increase. And they perceive over time that this number is probably going to live much higher and next goal is sort of three quarters, and to maintain going from there.
And the combination, it really accelerates their progress in infrastructure. So from Cavium really what they got was a nibble start-up mentality. You are going to hear from one of the co-founders of the company, Raghib Hussain later today. He is going to talk about their networking industry and the opportunities there. They got very valuable technologies from Cavium, leadership now in processors, leadership in security, leadership in accelerators. And so, they besides brought a very tenacious presence in datacenter and carrier which they had been trying to build. When you combine that with Marvell power and enterprise, their end-to-end product development capabilities and the operational excellence of the company, it creates a very unique value prop for their customers. They really are getting behind this combined company strategy. And in fact, I would screech from the customer point of view, the flat of engagement that we've seen since they closed the transaction has been significantly higher than before either company had ever experienced. I live of value quite frankly disburse a lot of time on the road. I beget personally met with the chief decision makers and CEOs and CxOs of the top infrastructure companies in the world across plenary of their segments.
I beget been diligent and I'll just give you one instance of a story. Three weeks ago, Raghib and I were - had the haphazard to present to the all leadership team of one of the top leading infrastructure OEMs in the world. And the meeting was really to contend a pretty significant multi-year agreement to accomplish multiple chips for this particular company and this would live a relatively a modern relationship or certainly an expanded one. And so after that meeting when they did the debrief, I asked the Cavium VP/GM who is now running that industry for us, I said, Look, I just got to ask, you guys beget gotten to this point on your own as a standalone company? And he said, "Well, accomplish you want me to give you the politically rectify answer, or the actual answer?"
So just give me the actual answer. And he said, "Absolutely not. There's just no way. They are viewed as being as too small, not having enough scale and not having plenary the capabilities." And so that opening now is one of many and you'll some more stories like this today of the power of this combination how that's really translating into significant modern opportunities because they are seen in a much different light now by their customers. And why accomplish they like? I live of value what accomplish they perceive when they spy at the Marvell portfolio that's combined, what they perceive is a leader, a leader in key technologies, in storage, networking, security, processors, connectivity. These are plenary the key elemental building blocks of what infrastructure companies are looking for. And it's not even that they beget a cramped piece participate here or there or a piece of IP.
We actually beget leadership positions from a technology perspective in each of these areas. And you are going to hear today from my team on not only where their leadership lies but what their strategy is to grow their position and continue to become a very compelling election for their customers. The combination besides with Cavium really creates a much more diverse company, which I reflect was one value that they saw in doing this combination. The first of which is from a customer point of view and I won't hook Tom's thunder, which you are going to hear more from him about the modern customers that they beget added and how actually while there is some overlap, there's a lot of incremental customers that they beget that Cavium didn't beget and vice versa. They besides now are a much stronger and a broader array of End markets.
As I mentioned, they really accept a much stronger presence from Cavium now in datacenter and in carrier. But the breath of the industry has besides increased too. And if I Go back to when I joined the company, I live of value probably the solitary biggest overhang that we've had and even continues to this day in some ways was the company's overall exposure to the HDD market. Classic Marvell if you Go back two or three years highly levered to HDD. That really created a cloud over the company in some respects. It always was a source of concern for investors. So look, we've taken a lot of pains on their own, right, to diversify their business. And you'll hear Dan talk about this more. With the combination of Cavium plus their own efforts to diversify within their storage segment, they beget now gotten their exposure today of HDD controllers that sell into notebooks which is really the heart of the concern.
The exposure now we've got down to about 7% of company total. That's as of today and they await that that number is going to dwindle as they head into the next year and the year after as they grow their industry and continue their pivot from consumer computing to cloud and infrastructure. I don't reflect that this pains that they effect in has been widely understood by investors. So I wanted to execute certain I called this out because I reflect that portray tremendous progress from where they were from a diversification standpoint.
Okay, so now let me talk about the portfolio and what are the elements of it. So, what's engaging is their businesses are quite diverse now. So, first, they beget established businesses in their company. This is really how they elude their company and how they segment their businesses with inside their portfolio. So the first is they beget established foundational businesses. These are businesses where they are the leader today. Two examples of these would live fiber channel adaptors which they got from Cavium as well as their HDD business. These businesses are strong. They are stable. They are profitable. They invest and manage them to maximize their leadership and their profitability.
So that's the base layer. Then they beget a number of growth businesses where either they are the leader or they are a very tenacious challenger. And examples of these comprise embedded processors, switches and PHYs, flash-based storage, SSD, and high performance Wi-Fi. Now these products are plenary in growing markets where they beget something that's differentiated and unique. They invest in these businesses significantly to grow them above the market and every one of them has a goal to live the market leader in their respective segment.
And then, finally, they besides beget several areas of strategic investment. These are modern bets that they are making. And these comprise automotive -- you are not going to hear about these today, these comprise automotive Ethernet, these comprise security solutions for the cloud, which they muster Liquid Security, which is a technology they got from Cavium, and besides this includes their server processor investment based on ARM. So these businesses typically they leverage IP they beget already got inside the company, but they leverage them into adjacent markets, so they accept a lot of reuse and benefit from that point of view. Every one of these has a lead customer or customers; that's their teaching customer, that's their sponsor, that's pulling us through to execute certain that they define the product correctly. They invest in these businesses carefully. They track them. They milestone them, and they execute certain that they are resourced properly to win. And so, when you step back, and you spy at the portfolio of businesses they beget inside the company, it's quite broad and it's quite diversified from an investment profile, and they reflect the combination of plenary these will result in profitable long-term growth with potential upside on the modern bets.
So, another considerable consideration for the combination was scale, and I reflect more and more today, you are not hearing this from Marvell, but you are hearing this from others that for digital semiconductor companies, scale today is incredibly important. It's no covert that the cost of developing these advanced node technologies is going up fairly dramatically as companies plod to more and more advanced process nodes, and obviously you are doing it to enlarge performance, lower power, optimize dye sizes, plenary kinds of benefits of making these node jumps, but they are getting more and more daunting by the day. I reflect this is a reality facing pretty much every company if you going to live in advanced node where 16 nanometer, 12 nanometer, seven or below, it's a different world than before. And so, from their point of view, they reflect that you probably exigency to develop about half a dozen chips or so at a minimum in a key process technology, just to beget spread the investment appropriately across plenary those devices. And so, if you don't beget enough scale, if you don't beget enough of a broad product line and you can't accomplish a significant number of modern products on a node, to accept your money back is going to live very, very difficult. So, this is putting pressure on a lot of companies, putting pressure on companies to scale up, it's actually putting pressure on their customers as well. Let me talk about that.
So, traditionally, their customers beget really had kind of two choices when they - tremendous systems companies, right, tremendous infrastructure companies, when they pick to -- that they exigency a key semiconductor component. One is they either want to buildup themselves, what they muster the plenary ASIC model, or you Go off the shelf and you buy merchant silicon, and typically this is where Marvell has been more in the merchant silicon side with limited on the build side. So build has made sense before, but it's getting to live increasingly challenging especially if you are a systems company, they are seeing now some of their customers having to staff teams literally in the thousands of people to execute this work, and obviously it's difficult to accomplish that many ICs if you are a solitary OEM. And so, that's a challenge for these companies today.
So they perceive a third industry model actually gaining favor. This is a model that Marvell has had for sometime, but they perceive it applying not just in their storage industry where it's traditionally set, but besides across actually other End markets. And this ally model is -- let me intricate a cramped bit more, is really the value proposition, look, everything in the gray that you perceive is IP that they are already developing today to spread across plenary the chips that they do, whether it's the CPU core, a remembrance controller, SerDes, plenary of these different blocks, they are developing these because they are going to disburse them primarily across plenary of their products. If you are a system OEM, you are just doing one ASIC, you beget got to Go either develop plenary this yourself or license it, it's very expensive and costly and it's time-consuming. And so, this model really allows us to accomplish what's in the gray and then the customer to really focus on what differentiates them, which is the red. That's their block. And this model they effect in space in their storage industry both in the HDD side and in enterprise SSD, it's been a space for about 15 years, it's been highly successful, because the customers are able to leverage and benefit from plenary the drudgery that they are doing, that's common across the entire company. And so, the benefit to the customer is obviously they accept proven battle-tested IP that's been in the market, to accept the benefit of their agreements with their suppliers, their cost structure with TSMC, the tools etcetera. So they can create their own unique thing, but they don't beget to accomplish the entire product. So it's really a win-win for both companies. And they perceive this is being a very compelling model especially as they beget combined.
Okay, so let me plod to the second participate of my talk today. So, probably I'm saying, "Okay, this is great, looks nice, it looks like you are building a nice company here. Everything sounds exciting. How are you going to grow?" I reflect that's the question that's on everybody's understanding today, "How are you going to grow?" So, Dan and Raghib are going to Go into the details, but let me give you my perspective. So the first is that the market forces that are out there are creating opportunities for us. There is a tenacious End market tailwind. The first thing, they can Go back to the last Analyst Day, is that there is an explosion of data that's being created and that besides needs to live stored, and there are zettabytes of data being created every year. It's doubling about every two years. This creates huge demands on the storage and network infrastructure.
The second is that the data that's being created at the edge more and more needs to accept processed, secured, and analyzed at the edge. And a lot of these applications now, they accomplish benefit from real-time decision-making where the data is occurring. And so, more and more they perceive this pull of the compute actually being pulled towards where the data is created. And so, that's a benefit to us. And Raghib will talk about that more.
The third is that in this more distributed world, you can't just secure the endpoints; you actually beget to beget a holistic security strategy to secure the entire data chain. And again, this is one where I reflect they can provide significant expertise here to provide robust security plenary the course from the datacenter, plenary the course to the edge through every point in the network. And finally, plenary of this has a major repercussion on overall power efficiency. It's probably one of the biggest antecedent of operating infrastructure today is simply the power bill. And this really plays to their strengths. Marvell, one of the hallmarks of this company has been their engineering expertise in developing low power SoCs, leveraging advanced process nodes and unique architectures. And so, when you spy at plenary of these combines, there are multiple tailwinds that are in their favor. And so, with that at their back, they reflect there are several very unique opportunities where they can execute an repercussion given that these market changes are happening, that are very specific to Marvell.
The first is in storage. So, plenary this data, as I mentioned, it needs to live stored somewhere, whether it's on cool storage, in the cloud on hard disks, or it's feverish storage on advanced glisten remembrance technologies. And so, their customers are looking for ways that they could accomplish this more economically and they could accomplish it faster. And so, as the leader in storage technology, they are in the middle of plenary of these major storage disruptions happening, because they are the core, they are controller, they are the brains, if you will, of many of the storage systems, and so, they beget unique insight and skill to actually influence and repercussion plenary of these tremendous disruptions that are going on in the storage industry.
The second is in networking. And they beget started seeing this last year, but there is a significant multi-year upgrade cycle that's occurring in the enterprise. And they disburse a lot of time at Marvell refreshing their own portfolio and optimizing it for enterprise over the last few years. They bring in Cavium strength, especially in embedded processors and enterprise, and they beget a very, very formidable portfolio in the enterprise, and if you spy just at their own results in their second quarter, their year-over-year growth just in Marvell networking was double-digits, it was like 16% year-over-year. So, they are already seeing the benefit of those modern products kicking in and the upgrade cycle kicking in. And they reflect that with the modern combined portfolio, they beget continued room for growth in the enterprise.
The third one is in the carrier market. And this may surprise you, but Marvell is going to live one of the most considerable companies to participate and enable the 5G rollout. Many of you were saying, "Marvell in 5G?" So you are going to hear more from Raghib on this today, but they beget very, very tenacious traction in this market, very unique solutions, and to live clear, they reflect that 5G is going to live one of the largest, if not the largest growth driver for the combined company over the next several years. They are very excited about this one.
And then finally, they beget emerging opportunities. These are some of their modern bets that they reflect provide tenacious potential for growth. The first one is in automotive. We've now gone to production with their automotive Ethernet products. You know, a year ago this wasn't even in their SAM, so we've made benign progress there. And besides in July, they introduced and took to production the first Xeon-class Intel-competitive ARM server CPU that's ever been introduced. There's been a lot of talk about this market for years, a lot of press releases; a lot of companies beget tried. But there's actually no company that's been able to insert and select to production a CPU with this flat of performance. And you're going to hear more on this from Raghib today.
Also, in HyperScale, we've had several public announcements now for their Liquid Security. You'll hear more about that today. But that's going to live a growth driver for us as well that's emerging. And each of these products is off to a considerable start. And I reflect when you layer plenary these in, whether it's store, the enterprise trends 5G, and these emerging opportunities in datacenter and automotive, they're significant. So let's translate that into the dollar amount, what's the total market opportunity. So at the last investor day, this was for the standalone company, they said total market for Marvell was $8 billion. When they announced Cavium they said that the SAM for the combined company was going to double to about $16 billion.
And now, we've got the team under one roof, we've been integrating, we've been looking at their opportunities, they went through their all portfolio review, and I'd screech the evolution has been -- they reflect there's about another $2 billion of market that we're now participating in that primarily is driven, as we've sort of gotten their arms around everything. One is the incremental 5G opening they reflect is pretty big. It's going to enlarge their SAM today. And besides we've now, because we've made such progress and we're in production, we've moved the automotive Ethernet SAM that was not there before into their SAM. And so that's a today number, that's $18 billion, which is pretty significant given the size of their company today.
So when you rupture the $18 billion down, $3.5 billion today is in storage. This is a industry where we're the leader today; they beget almost half of this market. This is a stable business. It's profitable, it's growing modestly. But they beget tenacious participate and a tenacious position, and opening to grow. After the combination with Cavium, networking now represents over $10 billion of SAM today. And this market is growing at 9% a year. They accomplish beget a tenacious position here as well, but they await to gain participate and grow faster than the market. Overall, these two markets combined, they're huge. This is $14 billion going to $17 billion just for the storage and the networking portion.
Now, as I mentioned, we're besides investing in ARM server, and they decided to rupture this out to live super limpid about where the drivers of their SAM are coming from. This is one where they perceive the addressable portion of the ARM server market today, at about $4 billion. That SAM they reflect is growing very fast, by about 14% a year. And so when you add that opening on top, which again is modern and emerging, total company SAM goes from about $18 billion to nearby to $24 billion over the next few years. So there's a significant opening if you spy at the evolution of where they were, screech at last analyst day looking at an $8 billion opportunity, to now just a yoke of years from now being able to address something like $24 billion of market. So you could perceive this market we're going after, it's healthy. It's large, it's growing, and we're very well positioned to capitalize on this opportunity.
So let me close. So in summary, the first point is we're planning to grow the company. We're planning to grow the company to live a leader in the infrastructure market. They beget the scale, they beget a diversified industry model, and that's one that's positioned to deliver long-term success. They continue to innovate and invest in the future, and this is going to enable us to allow their customers to disrupt their markets. One instance is 5G that you'll hear about. I'm besides especially proud that their team has consistently delivered and established a tenacious track record of execution that's going to live very considerable as they head into their next phase of growth. And so look, in short, with the team we've got, markets that they are going after, tenacious customer pull; I strongly believe that Marvell is going to drive ourselves forward. Their goal is to create a considerable company with considerable technology to enable the infrastructure of the future.
Thank you very much. Dan?
All right. Thank you, Matt. plenary right, so I'll insert myself first. I am Dan Christman. I am the Executive Vice President in Marvell, In-charge of Storage. Today, I am going to talk to you about their storage business. And they will talk about market dynamics, the opening in front of us as well as innovation and how they are targeting their investments. And most, importantly, I reflect their pivot to datacenter, infrastructure and really as a solutions-based storage company.
So, Marvell is the leader in storage and they actually beget significant participate here. As Matt mentioned almost 50% of the market is a market participate now for Marvell. We're investing in leadership which for their storage primarily means datacenter. They beget thought leadership. They beget a company that only has 20 years plus in storage can provide to the market that allows us to innovate, bring modern architecture to the market, and unravel their customer's toughest storage problems.
We are targeting growth segments which are helping us expand their SAM in the storage space and they are providing higher value solutions in the future. So let's talk a cramped bit about this. As you combine Marvell's classic storage industry of HDDs and SSDs along with Cavium's now fiber channel business, you actually accept the largest and broadest portfolio of storage solutions in the industry.
Fiber channel is over a $500 million in opening and Marvell is number one in fiber channel adapters. HDD solutions is $1.9 billion opportunity. Marvell is number one in HDD controllers. And they started shipping in high volume this year preamplifiers. Their glisten solutions business, which includes their SSD controllers, is a $1.1 billion opening today and it's growing fast.
Marvell is number one in merchant SSD controllers. And now they are affecting beyond controllers. They are providing innovative modern architectures and solutions. And they will talk about those in these slides. So as Matt mentioned, the storage market for Marvell is a $3.5 billion market today. It's growing at about 3% a year to $3.9 billion. Now if you rupture down a little, you will perceive the PC space is shrinking at about 5%. They perceive this is an district that's harder to differentiate, is less innovation, is less exigency for modern functions and features.
But the Edge in other market which includes automotive, industrial, video surveillance, home gateways, gaming, direct attached storage, this is actually a very highly resilient and stable market. Talking about 1% growth over the next three years, but really the opening is no surprise based on Matt's intro is datacenter. This market is growing at 9% a year. It's getting bigger.
We recognized this early and they focused their R&D dollars into this market. So, let's spy at their revenue here. So if you spy at the classic Marvell storage revenue, which again was HDD and SSD, you can perceive a yoke years ago almost half their revenue was coming from PCs. And you fast-forward it today and you'll perceive a much more balanced portfolio in the storage industry for Marvell. But a yoke of years out looking just at classic Marvell, you perceive the PC drops below quarter of their revenue while the Edge and other remains very stable across plenary three of these snapshots in time. The datacenter becomes very meaningful for Marvell. It actually becomes their largest segment in just a yoke of years. When roll in the fiber channel industry from Cavium and you perceive that Marvell has become a genuine infrastructure storage powerhouse.
PC is now down to 18% of their storage revenue and in a smaller percentage of the Marvell company revenue. So when they talk about fiber channel, today again it's a $500+ million opportunity. Marvell is the market participate leader with tenacious incumbency. They are the preferred supplier for leading OEMs and Fortune 1000 companies, and due to the captious and the sensitive nature of the data that gets moved in security onto fiber channel infrastructure, they perceive this is a very stable market with longevity. And the fact is that Marvell is silent innovating here, privilege we're helping extend their leadership through areas such as in-line security and NVMe over fiber channel and this is what customers custody about.
Now, there is no surprise, I reflect everyone here in this room understands that the HDD is going through a secular decline, privilege if you spy at units this is about a 10% a year unit decline in the market but due to the benign mix, the addition of preamps, the higher capacity drives going into datacenter, the SAM itself they projected decline about 7% a year.
Let's dig a cramped deeper, you perceive that really PC is the market that's most impacted really driven by the replacement cycle of SSDs in the PCs replacing HDDs. If you spy in fact at both desktop and notebook both markets are declining in the mid-20s. But when your plod PCs, you perceive the stable piece underneath which is actually relatively flat and if you dig in deeper, there you perceive really what Marvell's focused on which is a near line segment in hard disk drives, these are the cool storage drives in the datacenter where they manage tremendous Data, if you select a photo about a week later, it's probably on multiple cool storage HDD drives and in the datacenter.
In fact in calendar year 21 about 40% in the entire HDD market will live in the near line segment, this is where cost per gigabyte matters and without hard disk drives in a datacenter they would not live able to store their data. Why Marvell is focused here is number one it's a growing market, that's pretty obvious but number two we're very well situated to win this space, right, where we've been in this industry for over 20 years. They are in 21th generation now of read channel development, we're a technology leader in plenary the considerable aspects that execute these drives work.
If you spy at the aerial density increases that really drive the capacity, you spy at modern technologies, they talk about energy assist last time they hear, HAMR, MAMR, Dual Actuator, multi-actuator, these plenary select investment, they select a ally that understands how to accomplish this and Marvell is number one in this space for a reason, they project that we're going to continue to grow in this space with the market and even beyond the market.
I besides talked about in the nascence the fact we're going to live more of a solutions provider, so an HDD that means preamplifiers. This is meaningful revenue in the future it's meaningful SAM for sure, this is a subset of the SAM I showed you on the previous coast not incremental but because Marvell has a tenacious position in the HDD controller space, we're getting extremely tenacious pull from their customers to drudgery with them in the preamplifier space. They want partners who are actively investing with them in the HDD space and if you spy at the current wave, that's driving preamplifiers today, it's really about capacity increases due to adding platters in the drives.
So to expand capacity, they add more and more platters in these drives today and when you add a platter, every platter is two channels of preamplifier. So as you Go from two to three to four to up to 10 platters per system, you Go from one or two preamplifier channels to up to 20. So that actually becomes very meaningful where the content for a pre-amplifier in an near line drive can almost equal the content for an HDD controller.
Okay. So it's very, very meaningful. The next wave is really through technology innovation. Again as I talk about these acronyms HAMR and MAMR, these technologies that the customers are developing acquires Marvell that besides develop modern technologies, the controller in the preamplifier beget to communicate together to control these modern technologies, you beget to control the energy elements within the system with the preamplifier.
So these are modern opportunities for Marvell, they add more value and accept more content in these modern hard drives and what I'm proud to screech is that today Marvell is shipping preamplifiers into the market, we've qualified with their customers, their customers are shipping drives in the market with Marvell preamps and they await this to live meaningful, meaningful revenue next year for Marvell.
So the overall repercussion of the market dynamics I just described to you as well as strategic R&D investments really expose here the fact is that the notebook exposure is decreasing meaningfully in the HDD space for us, it'll live less than 15% in a yoke of years.
Our investment in datacenter is clearly paying off as you can perceive in this chart and they believe by diversifying their HDD revenue and by growing in preamps that they can partially offset this secular decline in the market and effect better than the overall market.
So let's plod on to glisten solutions. Now FMS which is a very distinguished expose for the glisten industry, it's glisten remembrance Summit, it's held in Santa Clara every year, it was last held in August, it was really a coming out party for Marvell, this is their booth here in the show, they came out as a company, they said spy we're more than an SSD controller company. Right, they are a glisten solutions company, we're focused on effectively or more efficiently managing the glisten based storage systems, this evolutions been driven by their skill to leverage the broader Marvell IP to enable modern profile factors for their customers, modern industry models and provide unique and innovative architectures at the platform level. This of course adds up for more content for Marvell and it grows their SAM.
So let's talk about now the evolution of glisten storage, if you reflect about it really started off as an HDD replacement, in the PC space people basically took a two and a half inch hard drive out, they swapped in a two and a half inch SSD drive, identical profile factor, identical interfaces, they went on to modern profile factors, they screech spy they don't know necessarily beget to beholding to this HDD profile factor, they can optimize for their PC. So they effect modern profile factors in that were smaller and more space efficient, they did modern interfaces like NVMe which took odds of the actual glisten and optimize the performance and after PCs went through this, the datacenters went through the exact identical cycle of replacement.
But now they're looking for more, they're looking for modern architectures, they're looking for modern industry models, they're looking for a platform based solutions. From a industry model in an architectural standpoint, I want to talk about this accomplish it yourself model, they did talk about it cramped bit last year but I want to talk more about it today, this is an opening that when you quiver off the limits of a hard disk drive, the mechanical limits of the profile factor limits and you screech spy I'm just focused on glisten memory, you really can now optimize, you can optimize for space and power, you can optimize for workloads, you can space the controller directly on your board or build an OEM and modern profile factor. This is really enabled by the fact that the customers can buy a controller from a company like Marvell directly.
They can source their NAND from multiple high attribute Tier 1 NAND vendors and they can build custom firmware. They really optimize this solution for their needs. Now Marvell's uniquely positioned here because we're a merchant supplier. They beget tenacious and long lasting relationships with plenary of the Tier 1 NAND vendors. They plenary drudgery together in strategic relationships as partners to develop these systems for their customers and then once we're on the board they can integrate additional functionality they perceive this is additional opening for Marvell, you can spy at the architecture can live changed. And this is basically more content and more value opening for Marvel in this space that this started in a datacenter but we've seen it plod beyond a datacenter now in Marvell's one designs already here you'll start to perceive revenue over the next yoke years here it's already a next year because it's kind of a paradigm shift for Marvell in the industry.
So let's talk now about platforms and how Marvell helping disrupt them platforms let me decode this coast first for you, you perceive this gray box here is the SSD controllers. They ship those today; will continue to ship those in the future but this red box is modern content for Marvell. I'm going to verse instance here this is an aggregator. Now their customers are looking to add more and more capacity as they add more capacity and they Go beyond PC centric profile factors to datacenter centric profile factors. The limitations start to become the SSD controller. The controllers can only ply so many NAND behind them and when you try to add more and more of these together you End up either with limited capacity or limited performance.
So Marvell's now insert modern aggregator, NVMe aggregator chips that basically even select in seamlessly stitched together multiple controllers or multiple SSD drives and present them as a single, high capacity, high performance drive enabling modern profile factors in a datacenter. In the second instance you hear a lot about Microsoft Project anally, open channel it's really about more efficiently managing the glisten storage at a flat above the drives which means are putting management kind of at the host. I know when you spy at doing that you screech well accomplish I want to effect that on my CPU and blow those cycles I could live renting out and the retort is always No.
And I reflect a Rag will talk about accelerators later but here they beget storage accelerators. And these accelerators basically will accomplish functions like compression, redundancy, security, IO virtualization and multi tenant systems and offload the CPU to assist more efficiently manage the glisten storage, it allow that cloud datacenter customers to rent out those CPU cycles. The last instance here is a revolutionary architecture that they actually announced to at glisten remembrance peak is an ether net bunch of glisten they muster it the eve off.
And this is an end-to-end chipset for Marvell that includes their controllers, it includes their NVMe converters and besides pulls in content from their networking group on the Ethernet switch side, if you spy at traditional server based storage in a datacenter when you want to add more storage you beget to add more compute they Go together, privilege in the Eve off even essentially this aggregated these, are you able to scale up your glisten storage independently creating a high performance rack of flash. Now this is connected over the Ethernet so it looks like its local, if you spy at latency is in performance it appears to live local to the host.
So when you want to add more drives and more capacity is simply add more drives into the rack, its scales linearly now they announced as I mentioned this that FMS has been very well received by their customers as an terrible lot of interest in this modern architecture. So the opening in the glisten market for Marvell is huge privilege if you spy at their classic controller industry this has a 17% CAGR. Their initial entry into here was in P.C. as I mentioned but we've pivoted towards a datacenter, in the datacenter customers value their performance their value reliability, they value advanced in modern features and architectures. This modern Sam on top here is actually growing faster than the overall controller industry so they End up with a 19% total CAGER for a glisten solutions business.
The gray box is incremental; it adds about 25% on top of their FI or calendar year 21 CAGR there or Sam. And they beget some industry here already it's relatively miniature but it's growing and going forward they talk about their SSD industry will really start talking about their glisten solutions industry with SSD as participate of that industry so you can perceive here how this plenary comes together. They recognize the opening in datacenter early. They aligned their R&D resources towards the datacenter. As they continue to execute on this pivot to live a provider of optimized solutions for the datacenter, you can perceive how this dramatically shifts their revenue profile.
These segments are becoming more and more meaningful. They started about a third of their business. Now it's half of their business. A yoke years from now it'll live three quarters of their business, so I'm going to summarize one more time for you. We're leading from a position of power in plenary of their storage businesses. Their shift to datacenter is well underway. We're seeing the results you saw the results today in their revenue mix. We're working with their partners of to pioneer modern and exciting innovative architectures to assist them more efficiently store and manage their data.
We've expanded from being a product solutions company to a more complete solution provider. And finally we're positioned for steady growth. Thank you very much.
Thanks, Dan. Guys, we're going to select about a 15 minute break, so let's live back here at 10.15 Eastern Time for folks on the webcast. Thanks.
Hey, folks. We're going to accept started, so if everybody can select their seats. plenary right, so we're going to accept the program going again. And it's my please to insert Raghib Hussain, who will select you through their networking business. Raghib?
Welcome everyone. It's benign to perceive so many close faces and the modern ones too. It has been a diligent year, plenary of excitement and a lot of potentials. I'm very excited to live participate of this modern combined company. And I'm fully committed to select it to considerable success. Let me expose why I'm excited. The potential that this combined company holds far exceeds what they had at Cavium. Now, just to give you a background, I'm founder of Cavium. And at Marvell I am Chief Strategy Officer and running the networking business. The scale that they beget in this combined company, the breadth of product portfolio, and the engineering knowhow, it is just incredible.
What excites me most is the team, the technology, but most importantly the innovative products that they are working on. And I'm going to participate plenary those detail with you today. But to originate with, they are tenacious in enterprise, and growing. They are very well positioned to live the leading semiconductor supplier for 5G rollout. They are driving transformation in compute for datacenter, and they are enabling the next generation of edge computing. So before they spy forward, let's start by taking a spy what is going to drive the growth in infrastructure, the spending in the next generation in the global markets. Exponential enlarge in devices is generating a massive amount of data. This data needs to live processed to generate value.
Now, engaging characteristics about this data, that it is perishable, it means the sooner you extract the value higher the value is. In a traditional architecture this data was generally generated by devices and brought to some central datacenter for processing, muster it a cloud. Now, because of the sheer volume of this data it is not feasible, and in many cases impractical to bring this data to some central devices -- central datacenter. And a great percentage of this data is used by the application which is captious in nature, for example, although the amount of data in 5G has increased, but the latency requirement remains the same, one millisecond.
The data generated by the sensor around the car has to live processed instantaneously for car to execute captious decision, like it has to apply the breaks. Now imagine if plenary these data was suppositious to Go to some central cloud for processing, it would beget been disastrous situation. So the network is morphing out of necessity and it is becoming more and more distributed. In other words, they will beget datacenter not only in the cloud but besides in the enterprise, in the carrier, and besides at many of the edge devices. So if you really spy at it data has got gravity. So instead of pulling data towards the compute, data is pulling compute towards itself. plenary these trends are massive opening for Marvell.
And not only compute has to plod towards data, it has to live efficient and optimized for the real-time application. It means that they exigency high performance compute and efficient processing at every node of the network, from datacenter to carrier to edge. And at each node they beget very specific requirements in terms of cost, power, and performance. It means one-size-fit-all is not applicable anymore. And with this distributed processing model the security has to live implemented at every node of the network. plenary these requirements are being addressed by purpose-built SoCs, application-specific hardware accelerators, and in some specific cases FPGAs and GPUs.
Now, both efficient compute and security play in Marvell's strength. The combined company has a comprehensive array of products, both for processors and networking. They beget a complete portfolio of processors ranging from baseband processor, to security processor, to multi-core general purpose processor, as well as plenary the course to ARM-based server processors. They beget complete Ethernet networking solution, from switches to PHYs, to NIC adaptors. And they beget high-performance Wi-Fi connectivity solution for both access point as well as client. So in other words, the depth and breadth of the product portfolio that they got is second to none. And plenary these products are going to drive the growth for their company in the infrastructure market.
We are using the portfolio to disrupt infrastructure market end-to-end, really enabling their infrastructure customer to accept the most out of this data economy. If you spy at these infrastructure applications they beget a lot in common. They plenary exigency high-performance compute, they plenary exigency security, they plenary exigency high bandwidth connectivity, and then plenary exigency low power efficiency. In many cases they are addressing the needs of these markets through a solitary piece of silicon, for example, the switches, the PHYs, and the multi-core processor. In some instance they are actually building market-specific application-specific optimized solution using their common portfolio of IP. For example, baseband processor. In other words, these markets beget common characteristics and they are leveraging their investments across their infrastructure markets.
Our product portfolio has a great growing addressable market. They beget established here in a base $10.5 billion SAM, which is growing at a 9% CAGR. Processors and networking are growing faster than Wi-Fi. In addition, they beget $4 billion SAM for ARM server processor, which is growing at a higher CAGR of about 14%. So, plenary in all, their SAM CAGR is 11%. Here is another view to spy at their base $10.5 billion SAM mapped to their target ends market. As you can see, it is fairly distributed across plenary four market segments. As a company their IP, their R&D, and product portfolio are well aligned with the major market trends in the infrastructure.
So let's talk about enterprise. Both Cavium and Marvell has a tenacious position in enterprise. It is about $2 billion market, and they beget about 30% share. Enterprise SAM growth for us is higher than the overall market growth, and it is because we're expanding their market participate through modern design wins. last time, in 2017 investor day, they talk about upcoming enterprise upgrade cycle and how Marvell is positioned to support upgrade from a gigabit to a multi-gigabit driven by the bandwidth needs. As you can perceive from this chart, the IT upgrade cycle is here. IT budgets are expected to grow driven by the needs of either upgrading the outdated equipments or by the security concern. It is just getting started. And consistent with these cycles of the deployment of the infrastructure equipments, upgrade equipments, they await it to continue.
While other companies beget lost their focus on enterprise, Marvell has invested in innovation in enterprise, building targeted solution with feather that their customer wants. It is silent a multibillion dollar market, and their OEM customer needs product to enable solutions for their End customers. They exigency the latest geometry node, they exigency the low power; they exigency the features needed for the evolving requirements on this industry. By serving the needs of their customers Marvell is already growing its participate in enterprise. last investor day they talk about 25 modern product, at that time they were ramping in revenue. Today, the revenue generated by those products is about $200 million, and it is silent growing.
With the combination of Cavium, now they beget a complete product portfolio from access to aggregation to core. And that will continue to drive their participate in enterprise. Their merger brought together a complementary power that enable us to provide complete solution for their customers. Cavium was tenacious in aggregation and core, and really the processors, and Marvell is tenacious in access and switching networking. Together they are able to provide complete solutions, complete platform for their customer from access to aggregation to core. And that makes us the privilege strategic ally for their customers. Their complementary customer base is besides a tremendous value for us. For example, Cavium had significant presence in some large-carrier OEM as well as great server OEMs, where Marvell did not beget much presence.
Now, with the combination of the company and the combination of the product portfolio, now they are considered a strategic partner. As a result, they are getting networking -- switching design wins besides in those End customers. So let's select an instance of a typical security networking appliance out there. As you can see, that there are always a switch and a PHY setting on the motherboard along with the processor. Processor decision are generally made first. Before, neither company had the complete solution. Cavium had benign established presence in the processor, Marvell had established presence in the switch and PHY. Now this picture is a coincidence that they both were in the identical boat; however in many designs they had some third-party vendor serving the other side.
Now, with the combined portfolio they are able to provide the complete solution for their customer. And along with the processor, they can actually provide the switch and PHY solution as well so that their customer can execute their decision upfront to bring the privilege solution. They beget already started winning designs in this area, and there are plenty to go. This is another example, Marvell is tenacious in switch and PHY but they did not beget many core processor to address the needs of aggregation and core. Now with the combined company, they beget complete platform with a switch, and PHY and a processor and the platform solution serving the needs of their customer from access to plenary the course to core.
So you can perceive the combined IP and the product portfolio that Marvell and Cavium bring together uniquely position us for tenacious growth in enterprise. On top of that, they are in the upgrade cycle, it has just started. And they beget the flush product portfolio with the latest feature to maintain driving it -- to continue driving it. Due to their commitment with enterprise their customers are considering us a strategic partner. And that is a position of power for us which will continue their growth in enterprise market.
Let's plod on to datacenter. As modern compute models are established they beget multiple high-growth opportunities in datacenter. Cavium has had a tenacious presence in datacenter through security and networking services offload. It is about $2.5 billion market, and they beget about 10% share. In addition, the ARM server processor in datacenter has about $4 billion SAM. We'll contend more details about it in subsequent slides. They plenary know datacenter compute is changing driven by multiple trends. The first one is distributed security and network services. As the datacenter is evolving, driven by the needs of elasticity and virtualization, the network services are being implemented at every node.
The second trend is cloud-optimized ARM server processor. And then they plenary know there's a modern trend, simulated intelligence, and they will contend more about it. Marvell has been market leader in providing efficient compute security and network services offload for over a decade. If you spy at any enterprise security or network appliance, and if you open it up, you will perceive that OCTEON and their NITROX processor are in it. When it comes to security and the data plane processing Cavium has been market leader for over a decade. In cloud, security and network processing requirements are changing and it is getting distributed as they talk about. And as a result, these are implemented in every node, but there besides you exigency the similar character of acceleration. Marvell's Liquid Security product lines are designed and very well positioned for that market.
We are engaged with plenary the HyperScale datacenter providers, as well as several data platform companies. So two of the HyperScale beget already announced their security services based on Liquid Security. And they are engaged with many more, so you await to hear more about this. This industry is in early stages privilege now, but it's already generating a benign revenue, and has a significant growth potential.
Server for datacenter is a huge opportunity. It is about $16 billion TAM, and they await that ARM servers can address about $4 billion TAM. Marvell Thunder 2X is the first Xeon-class processor. When I screech Xeon-class processor, it is really the dual-socket ARM server processor which has the performance as well as remembrance bandwidth and accelerate and connectivity of a really Xeon class which can live used in a general purpose server application. They beget wider software and hardware ecosystem. ThunderX2 platform has gone in production in July, and they are engaged with several HyperScale End customer at various stages at EVT, DVT, and qualification and application tuning.
We are working closely with several vendors in U.S. and Asia. If you select a spy at these recent announcement by their customers, several customers beget announced platform based on ThunderX2, and then there are several independent third-party analysts beget published the benchmark comparing ThunderX2 with the Intel and AMD processors. One, of the -- one which is Astra, which is the first world petascale supercomputer, it is among the top 100 supercomputer in the world and it is based on ARM server processor ThunderX2. It has 145,000 processor cores, ARM cores, delevering about a 2.3 petaflops of performance.
Now, one of the reputed analysts is AnandTech and this is what he has to screech about ThunderX2, "In short, ThunderX2 is the first SoC that is able to compete with Intel and AMD in the general purpose server CPU market. And that is a pleasant surprise. At last, an ARM server solution that delivers." They are seeing a benign traction in this market, and they await it to live long-term growth driver for Marvell.
Now, about simulated intelligence, they plenary know simulated intelligence is the next -- is the modern gold rush out there. It works just like their brain. So for example, the fact that they know this is a bottle because their brain neural network has been trained over time that things that spy like this is a bottle, by different types of bottle, the perfume bottle, the wine bottle, and so on and so forth. So that participate of the neural network is called training and learning. And then when they perceive something that spy like a bottle their brain predicts that it is probably a bottle, that participate is called prediction or inference.
Now, training is generally done in cloud because it is a leisurely process, it's a batch process, it requires a lot of data and it does not exigency to live in the real-time. It can be, it can live done in a batch process way. However, inference, it's not only done in the application in the cloud but besides in application enterprise carrier and the edge in the edge devices. Now inference has to live done in the actual time and instantaneously because this is where you're predicting, you cannot select maintain the consumers waiting or user waiting for the result, right. So as a result of that, it has to live done at every application. As they plenary know one vendor out there has made a fortune out of exploiting training. However inference today is generally done in the software because the number of applications that are using inference is in a growing stage at the moment.
We believe that inference is going to live much bigger market in the overall simulated intelligence and it has a long term growth potential. Inference requires a purpose built solution optimized for a scale and power and cost efficiency, plenary of these AI processing plays in the core power of their company, they beget a DNA of multi-core processing, hardware acceleration flowing out, engine scalable architecture. Gavin was working on AI for the last several years and they beget developed some core IP and architecture.
Now what they are doing now, they are building a purpose-built inference processor. Size properly, for volume application, application of scales, so that it can plug in every server and every edge devices. That is a low-power that is a programmable solution to adjust the evolving neural networks needs. And it is besides pluggable through the existing software ecosystem. They are actively engaged with several hyperscale customers out there and co-developing it with one lead partner. They beget been working besides closely with the ecosystem, you must beget seen intelligence related to glow compiler initiative driven by Facebook. It is an initiative to really standardize the inference usage. They perceive AI a multi-billion dollar market opening for us and a captious office in future Marvell products.
Well, they are very excited about the prospects of this. They are not adding it in their SAM at the moment. However, we'll maintain you updated with the progress that they make. So let's select a spy at Edge and other related areas, the Edge and other district for us is really the automotive, the industrial, the video surveillance, the home gateways, gaming et cetera. It is about $2 billion market and they beget about 14% share. If you select a spy at plenary these products, Marvell has been present in plenary these products for many years, a major result of the data economy is that the trend that compute is affecting towards the edge, towards plenary devices.
As a result of that, many of these devices are becoming very sophisticated, in some cases really becoming a mini datacenter like for instance in car. Marvell is actively engaged with plenary of these trends, the automotive market is going through a massive transformation, traditionally in car electronics was connected together with the low bandwidth interconnect. With the introduction of advanced driver assist and ultimately the autonomous cars, massive amount of data is being generated by the sensors around the car and it needs to live transported and moved around in the car at a quick pace. This requires touchstone base high bandwidth networking, Marvell is leading this trend, Marvel has long history in automotive industry, it is a high barrier to entry industry.
It requires a specific attribute as well as supply chain requirement. Over time, they beget established ourselves a credible automotive supplier. In 2017, they introduced the first secure networking sites and five product for automotive working. This enables the data by the sensors to live moved around in the car at a gigabit accelerate and car can execute sense the wall around it and execute a actual time decision. They are one of the early leaders with design wins and many Tier 1 OEMs. This design takes time but they are a significant long-term growth potential for Marvell. And this year they beget included this SAM in their overall SAM and they perceive it about half a billion dollar opening and this is growing at a quick pace.
Moving on to carrier, this is a considerable growth opening for Marvell and it is a market they are extremely excited about, 5G is here and it is happening and Marvell is well positioned to live the leading silicon supplier for 5G. Carrier is a massive SAM for us, it is about $3.5 billion SAM market and they beget about 10% participate which means they beget a lot of room to grow. Their carrier SAM CAGR is higher than the market growth, this is mainly because in the 5G platform, they are increasing their content and hence increasing overall portion in the TAM. building on a decade of innovation in 3G and 4G, now they are established to live a tenacious leader in 5G, here they are positioned to disrupt the market and grow their share, I'm really excited about 5G and I perceive 5G as the biggest growth driver for this company, the combined company has the broadest IP portfolio and capability is needed to enable to serve the requirements of the infrastructure market.
In fact, they are the leading market silicon supplier with end-to-end capabilities. From DSP, baseband processing for ARM multi-core processor, for control and data plan, for ARM SoCs, from security, Ethernet connectivity as well as software for the complete solution. They not only beget plenary the captious building blocks but over time, they beget established ourselves a credible supplier to deliver high performance platform for baseband applications. This makes us a very attractive silicon ally for the carrier OEM. While today they will talk a lot about wireless and the base station, it is considerable to note that the combined company has much broader presence in carrier. Their products are designed in multiple appliances for both wired and wireless side of the carrier network. Their position in base station has grown with every successive generation of wireless infrastructure deployment.
When they engage first time carrier OEM came to us, at that time they were security leaders, so they came to us to provide a solution to secure the link between base station and the core. However with their multi-core capabilities, they were able to provide a solution for protocol processing for transport in addition to security. So in 3G, they were solution for the transport. Working closely with a lead partner, they were able to develop baseband capabilities and 4G they were able to tender baseband processing in addition to transport processing.
I'm proud to situation that today the base station built on their products for 4G are being deployed across the world. And specifically the LTE network of a region with over billion population is powered by their baseband processors, they beget shipped over seven million base station processors as of today. And now with the combined company, if you spy at the requirements of the 5G base station 5G deployment, it has a requirement of low latency, high performance compute and high performance capability of security, it plenary aligns with Marvell's core capability.
So as a result, they are able to provide the complete 5G platform, looking at it another way, this is the complete stacks of the base station, in 3G they were able to provide the protocol processing, in 4G they expanded their offering to cover the baseband processing and now with 5G, they are going to tender the complete digital portion of the baseband processing. In other words, they are taking the workload, which were traditionally done in FPGA. If you spy at the 4G base station, in the main card they used to beget Octeon processor and in the line card, they used to beget three baseband processors.
Now in 5G with the key requirement, they beget two Octeon processor in the base main card and typical configuration of 5G has two line card and each one beget not only a three baseband processor but besides Octeon processor. Now just like in enterprise playing through the combined portfolio, they besides beget a switch and PHYs in this base station application. Translating it into what they plenary custody about in 3G, they had a content of yoke of hundred dollar in the base station, in 4G they increased by three to four times and 5G they are going to enlarge it another 4X. It means that the base station shipped by their OEM partners is going to beget the content which is quadruple in 5G compared to 4G. But there's more, if you spy at the number of OEM providers in 3G timeframe, carrier had a election to pick from a great number, typically carrier pick three OEM for a specific region.
In the 4G timeframe because of consolidation, it was reduced to seven, eight and now it is really reduced to only five OEM providers out there. And now because of geopolitical situation, several great countries of the world are limited to election of only three OEM providers as they are engaged with plenary of them. So if you select a spy at this chart, it expose there is a lot of activity going on in 5G, there are many announcements related to 5G and it is really picking up. Initial deployment of 5G is going to betide in U.S., China, Korea, Japan and India and of course the comfort of the countries will succeed from here.
Carrier infrastructure deployment cycles are long, it's a long-term business, if you select a spy at let's screech for instance 3G it expand over a decade, the carrier are silent OEM are silent shipping base station for 3G. 4G had a keen RAM driven by the bandwidth needs of the applications and it has had a benign run. They await 5G to beget similar RAM driven by the application needs of the various applications that are driving 5G deployment.
So base station based on the design wins and their attraction with the customers they await 5G to live the leading growth driver for this company because they beget proven track record and IP. They beget position with enlarge content and they beget broader traction with multiple OEMs, so if you want 5G in your portfolio guess what Marvell is your stock, so by now you must beget figured out why are we, so excited about the potential of this combined company.
We are tenacious in enterprise and growing significantly. They are investing in the privilege products to drive the growth in datacenter and Edge. They are well positioned to select the leading position, leading silicon supplier for a 5G roll out and they are leading that disruptive trends which is on server processor datacenter security, automotive networking and simulated intelligence due to driven consolidation in the semiconductor industry there not too many companies that are investing in the long term innovation and the growth drivers.
Marvell is a unique company which is not only established today and growing but besides committed to the long term growth. Thank you very much. I'd like to invite Tom now to give the how to drive the growth.
Good morning, everybody. I guess mic is on now. It's benign to perceive you plenary again, it's been a screech six quarters since they were last together. As you can perceive this is a very different company. Today than even six quarters ago when you guys were last here Dan, Raghib, Matt talked about a broad portfolio, focus on the infrastructure market a lot of attention from customers and you're probably plenary sitting there going yes, expected you to screech that today so what are the customers beget to screech about this and so that's why they asked me here today to kind of give you the perspective from the customer side of how this acquisition is being perceived and how this modern company has being perceived.
Okay, so as you recollect when I was here last time. It was March 2017 they were plenary in this room together it was snowing outside and I had been with the company for about three months and I told you plenary I said spy in order to accept this thing on track they first order a industry is to align the sales and marketing office with the strategy of the company at the time, okay and I said these are the things we've got to accomplish and what we've got to accomplish quickly, so what it would how we've done what they accomplish here, so I talked about a coverage map in the first thing I had to accomplish was build a team, so I rebuilt my entire staff and then they drove that plenary the course down to the bottom flat probably about half the people on my staff beget worked for me before they understand my system.
They understand what I want to accomplish they understand what I value. The other half beget not but having guys who drudgery for me before on there beget been a considerable value to the team in that they've helped him quite a bit. They talked about relationships, in an SOC semiconductor company today, it is more difficult than ever to execute the sale there are more stakeholders involved in every transaction than ever before so they actually developed relationship major cities across plenary the major customers they executed those major cities and we've expanded their relationships across plenary the major customers. They had to antiseptic up the channel. So they went down to a solitary global distributor they reduced plenary the distribution partners in the regions around the world.
We wanted to enlarge their scale with plenary these partners to accept more participate of mine, so they restructured the entire channel that was plenary done and they did that literally within the first three or four months of me being here. On the sales strategy side you know as I told you last time I'm very much metric space data driven, so they did a lot of drudgery around analytics metrics tools and processes understanding the selling capacity of this organization, collecting plenary the data and analyzing the life cycles of opportunities how these opportunities converted to revenue plenary that stuff drudgery started done started then they started collecting data they worked on the historical data they had and they effect plenary that in place.
We did the entire implement flow. They focused on solution selling and they really kept the organization hybrid to extend their reach with the variable expand sales favor of reps and distribution that they have. THEY talked about account penetration. They beget maintained plenary of their top accounts and as a matter of fact we've grown their top accounts from the Marvell side since they were last here. They grew some modern accounts, they added a number of nice accounts into the top ten list and as they integrated Cavium in. They really only had one customer in common in their in their in their respective top ten lists, so they had a very complimentary top ten list we'll talk a cramped bit more about that later but we've got a very benign account list going forward and we're continuing to drudgery their relationship matrix and expand their relationships up and down we'll talk a cramped bit about some of these actual meetings and a cramped bit. On the marketing side I told you they had fix digital marketing, the first thing they did is they spent a lot of time perfecting their digital marketing techniques, a lot of very targeted account based marketing programs.
Targeting customers with specific messages in a very cost efficient manner, driving into their website, cleaning up their website adding a lot more pertinent content to it plenary of this stuff has resulted in a lot more activity and a lot more modern customers coming to us. They basically looked at and examined their public relations and trade shows strategy making certain they got maximum revert on money spent there and they instituted quarterly reviews with the industry units to execute certain that they had a marketing cadence for modern product introductions and marketing activities every quarter so a lot of drudgery on the marketing side.
I talked to you about industry development and I told you that industry development in my world was two things creating preference and awareness at the End customer for Marvell products and building relationships with people in their industry that needed to that were considerable to us but weren't necessarily going to buy from us and so what accomplish they do, so the first thing they did was they built out their automotive BT BD office when I was up here last time you guys didn't even know they had an automotive industry within three months of being up here last time, they had introduced the world's first gigabit secure gateway for the cars and my sales organization was overwhelmed with the activity coming in and plenary the design activity that was happening with that product line.
We now beget complete coverage in plenary the car companies in the world with BD folks and so that's been built out. We've built out or are now processed to their participate participating in building out their service provider organization and we're cultivating a lot of strategic partners in the industry that they exigency in order to continue to Go forward, people like Nvidia with the automotive market, people like Intel they beget strategic relationships with as they Go forward, so they did plenary of this through 2017 as we're enter ending 2017. They decided their life was far too simple so they bought Cavium and they started it plenary over again and this is what we're actually doing now this is where they are with the Cavium integration. In November, they announced the transaction, in July they closed the transaction and in those eight months they spent massive amounts of time planning for the day one activities. They spent time coming through the organization making certain that they were going to assemble the best organization possible.
We spent time looking at plenary the implement flows to execute certain that they were going to select tools the best tools from both organizations and disburse them to create a best in class implement rush for the company. And so on July whatever it was when the transaction actually close, they immediately instituted a program. Today they beget the best of both worlds where the top talent for both companies 62% of the organization came from Marvell, 38% came from Cavium a very benign blend of talent across the sales organization. Literally within weeks they brought everybody into Santa Clara from the Americas organization sales and FA's and they did product cross training to accept these guys affecting quickly on selling the modern merged portfolio.
And they did a recording of plenary that so that EMEA and Asia at least until they can accept out there doing things in person had training as well and they could actually participate online tools and systems are merging now, will live actually introducing the merged system within literally within a yoke weeks and for those of you who know me in my world, in the world of SoC semiconductors I view value as design wins, design wins are the lifeblood of this company and so starting in the next fiscal year everybody goes on the identical pay draw where they will compensate for value creation in value created is design wins are really focused on keeping the keeping the life blood flowing getting into production and then driving it over to Andy and the guys to execute it on a regular basis.
And then after plenary this was done I heard I hauled Murphy plenary over the world and they talked to plenary of the customers about this company and what they were actually doing and so how did they react. Okay, so they basically told the tale of their observation of the market what they saw what you saw in the video today what you've heard from Raghib, what you heard from Dan about market dynamics that were driving the strategy that we're deploying and they talked about the fact that they had a focused infrastructure powerhouse that they were building a pragmatic undergo ally that's going to select their intellectual property portfolio and unravel their problems with everything from IP based semi custom designs plenary the course to touchstone products and we're going to apply this IP, this IP to their next generation learning machines and every customer universally to a customer yes they accord with plenary of your observations in the market.
Yes, they got slides just like that yes they were besides extremely surprised by the breadth of the intellectual property and every one of them to a customer acknowledged a wish for a deeper and broader relationship with the company and so they really told the, there are their customers and they are out there we're building a company that thinks ahead so you can too and you heard that in the video today very well received by the customers. Now you know I'm going to elude the risk of offending the animal rights activists in the room again. I showed this to you guys last time and I basically said spy if you want to trap tremendous fish, you've got to fish where the tremendous fish are in other words we've got to deploy their sales and marketing organization, align their resources with their largest customers. This is the best revert of expended sales and marketing money that I can deliver to the company.
We beget to win the largest customers in their chosen markets first, the comfort are going to succeed so they are aligning the sales organization we're deploying their resources next to the largest customers in every market. We're going to disburse the broad portfolio in a collaborative way; we're going to unravel their customer's problems. They beget actually built the company the customers like and want to accomplish industry with. So where are these tremendous fish, who are they what are they doing so privilege now as they merge their customer list as they merge the companies eight of their top 10 customers are currently over $100 million, the next two are within spitting distance and we've got a number of customers today who are sort of mid Tier customers that you heard about from Raghib and Dan that are probably going to push their course into that top 10 over the next yoke years.
All the top accounts are direct but we're silent investing in the channel, things like putting deploying an FAE certification program this past year flat one flat two, so that FA's are tested literally given a written test basic lore flat one, basic advanced lore to support products flat two and flat two actually has monetary repercussions for the distributor, so they're motivated to continue to add support and add resources to their product line. So select a spy at this list Enterprise carrier datacenter it's a who's who of blue chip customer list if you want to accomplish industry in these markets and these are not aspirational customers, we're doing industry with every one of these guys today they are contacting connected with every one of these guys today.
On the edge side I specifically left automotive in there to talk about, primarily because the car is the ultimate edge computer, it's a learning machine that's creating massive amounts of data it's making latency sensitive actual time decisions on that data. It's filtering that data it's sending it back to the core that current millions like it are doing that that data is being filtered at the core learning is taking space and it's going back to the edge the data gravity that that Raghib had actually explained to you. And so as they came into this thing with that gigabit gateway that they introduced last year, they started doing industry with everybody on the planet who you want to accomplish industry with, who is an automobile manufacturer design wins at major car companies that will start to actually generate revenue next year and relationships with every Tier 1 on the planet that they really custody about, so you can perceive a number of up there from if you Chrysler to Ford to G.M. to Bosh to B.M.W. We're working with every one of these companies privilege now.
So it's a very tenacious customer list okay, so we're really using this IP portfolio in this industry. To provide platform solutions using supple industry models to unparalleled support to live a valued ally to their customers. We're literally selling these technologies into these markets okay, so how is this playing at the customer's Let's select a spy I will Go through three examples of actual customer meetings on how this portfolio is creating tremendous amounts of opening from one side to the other okay. So this is a this is a datacenter platform if you reflect about a mega scale datacenter it's got a hardware layer, resource management layer, virtualization layer a services layer. They typically play in the hardware layer as you would expect.
Now, they recently had a meeting, Raghib and I with the CEO of a cloud -- public cloud provider, a hyper scale datacenter guy. And so, at that company before the merger, Marvell played in there in the storage and in the networking side. They had SSD design wins where they are working with them with their SSD controllers so that they can provide purpose bill feverish storage for their storage arrays. They are working with them on some of the datacenter glisten system solutions that Dan talked about, some of the aggregators and accelerators. They worked with them on switches and PHYs, and some of their two and four core Armada processors for data plane and control plane processor. So they had an established relationship with this company.
Cavium came in from the server side with server-based CPUs, network offload in the profile of smart NICs for Ethernet and security offload with the Liquid Security platform. So they were having this high-level meeting, the CEO of this company and his lieutenants were sitting there, and they are discussing the portfolio. Now, I beget to read some of the comments he made, because they were more astonishing to me as a sales guy, but I literally wrote them down. During this meeting, as they were going through them, the CEO said to us, "It is in their interest to enlarge the industry relationship with Marvell." simple sales guy, CEO sits there and says, "It's in their interest to enlarge the industry relationship with Marvell. This doesn't suck. This is a benign thing." Okay? "It's much better than you guys think. You can't execute me -- web paper bag, don't muster us, they will muster you."
So I literally wrote that one down, and they continued to drudgery with this guy, and they are talking about plenary of their products and how they map into his products, how they can accomplish more industry together. And he says, and I quote, "We exigency to beget this flat of allegiance around plenary things Marvell is doing." Again a quote that I kind of like as a sales guy. And so, they literally walked out of there. And the top lieutenants at this company are awake of the fact that this guy wants to basic beget a broader relationship with this modern expanded company, and they beget actually had more opportunities open up there in the past several weeks since that meeting that you can possibly imagine. Okay? So that's instance one.
So let's talk about carriers and in particular they will talk about cellular base station carriers, okay? So Raghib talked about this, talked about 4G, 5G; both companies had a history in this market from LTE and endpoints before. Marvell primarily selling switches and PHYs into this market and the two and four core processors for data plane, control plane mostly in backhaul applications. And Cavium was selling their baseband in plenary multi core integrated processors in there both of us had a history. So they were having a meeting again with the CEO of one of their customers who provides base stations into this market. And they were kind of -- he was kind of acknowledging, you know, they are sort of very captious to each other at this space as 5G market acknowledges. And he proposed that maybe they want to deem having a written agreement, a multi-year supplier agreement where they could maintain focused on each other. Again, as a sales guy, they fight tooth and nails to accept these multi-year supplier agreements, accept customers entrust for long-term. And they got a customer proposing that they sign this agreement so that they could wait focused on them. So, very, very tenacious relationship with these guys, this relationship is affecting forward in a very benign manner.
And finally, they will talk about enterprise. You spy at enterprise, and in enterprise you got access, aggregation, and core. The accesses were the -- the humans interfaced to the network aggregations where they plenary aggregated cores where plenary the routing and processing takes place. And so, they were meeting with a lot of -- they met with the execs of almost every customer who has portfolios to sell into this market, people like Cisco, HP, Dell, Aristo, Juniper, Extreme, Lenovo, these are guys who sell into this enterprise market, broad portfolios of equipment. And again, they were meeting with the CEO and several of his high-level executives. One of these companies provides a broad spectrum of products in there.
Now, Marvell has done industry in this market primarily through access going into aggregation with switches and PHYs, again those multi core processors for control plane and data plane, and with Wi-Fi for enterprise access point. And they pushed into aggregation. Cavium is coming from the other side doing industry primarily in the core with ARM-based CPUs and security processors and coming this way, so that they kind of met in the middle. And during this meeting, the CEO is acknowledging the broad portfolio and how captious they are to their industry going forward, and he asks Matt for a favor, "You know, I exigency you to accept this one product affecting a cramped bit. They got customer commitments. Can you assist me?" "Yeah, yeah, they will help," Matt made the commitment and went away. And as they are getting ready to leave, they said, well, you know, since they are asking each other for favors, they said, "Hey, Mr. CEO, you got a switch design that they are fighting for in one-year industry units. Could you let us know how they are doing there?" He said, "Well, I will check." And literally within two weeks they won that switch opportunity.
Now obviously, they are not going to win this opportunity, if they don't beget everything it takes to accomplish it technically and beget the pricing and the requirements of the customer needs but getting a cramped phone muster from the CEO aphorism hey how is Marvell doing in this switch opening socket, certainly doesn't damage when they are actually competing in there and so these types of things maintain happening, they maintain happening over and over again.
In the past, they were the guy with the narrower portfolio that would struggle to compete with the broader portfolio. They are now the guy with the broader portfolio and the customers are valuing it and so you could see, this is a faultless example. These three examples that I've given you are faultless examples of how portfolio breath is helping us and these are just three examples. This is happening every day, they are seeing people coming and screech this portfolio is very valuable to us. They want a broader relationship with you guys. Okay, so just to kind of wrap it plenary up, we've been on a journey, the first 11 months they took sales and marketing apart, they effect Humpty back together again.
And then, they did it plenary over again after the Cavium acquisition to give you a more predictable analytical metric driven function, okay and so they are largely done on the integration this thing is done. They created this infrastructure powerhouse for you and customers are sitting up and taking notice. We've had high-level meetings with pretty much every customer they want to accomplish industry with and everyone is excited about what they had to tender and how they can grow together. So as far as I'm concerned, their future is so bright, you got to wear shades. I was going to bring in the custom sunglasses for every one of you guys, modern 11 lens Marvell on the other, but Jean crop my budget again. I didn't beget the money to accomplish it. So best I can accomplish now, screech thank you for listening, and insert Jean, who is their CFO.
Thank you. Thank you, Tom, you know what, when you submit your revenue synergy plan, I'll wear sunglasses. So run up. So as you can see, it's a truly exciting time to live at Marvell. Their team talked about the tremendous opportunities they beget ahead of us and that they besides talked about how they build a complete infrastructure portfolio to address those opportunity. So well how accomplish they disburse the next minutes to tie together what you heard of this morning without monetary model?
First as a team that they reflect about building shareholder value is to really live with the powerful industry model to focus on infrastructure market to generate a top-line revenue growth earning expansion and besides returns to shareholders. But they reflect about is as their team talk about opportunities and their unique position. Matt talked about the characteristic of infrastructure market which minister to beget a very long product cycle, it requires a unique IP extraordinary engineering execution to drudgery with the customers. So the barrier to entry is very high and Raghib actually gave you real-time instance to talk about 3G, 4G and the 5G product cycle, they last 8 to 10 years.
And then, Dan talk about how they drudgery with the customers, create innovative solutions in the storage market. So when you hear plenary those that's like music to CFO's ears and I'm pretty certain you are plenary love it too because what it means is only build their monetary model, it's predictable revenue stream, high and the stable grievous margin and the long -- and the consistent of free cash rush with a higher terminal value. So those are the characteristic of infrastructure market, and besides means in their monetary model, they actually don't beget so many what if assumptions which are focused on retreat and parameters like NAND spot pricing next week. I can promise you that's the input in my model.
So I'll cover three topics. First, I'll accomplish a quick recap of their monetary performance. Secondly, I'll contend their long-term monetary model and underlying assumptions. Third; I'll talk about what accomplish you custody most is the capital return. Matt showed you this charter earlier about their monetary performance since the last Investor Day. I'll provide you a cramped bit more details to expose you as a company, how they beget fundamentally changed the structure of their industry model.
So just as a quick reminder, the all management team, joined Marvell after Q1 fiscal '17 and that time Marvell was a company that served a broader role for consumer and the market. The grievous margin was low 50s. It's actually quite consistent with the consumer semiconductor companies. The company besides invest in great amount of R&D in plenary different kind for fancy consumer and the markets. The investment is large, the product cycle is very sharp, it's a year and some of them never generated the top-line revenue growth and earnings on the investment. So if you spy at the operating margin back then, its barely single-digit.
So during the last two-and-a-half years, their team really pivoted the industry into the infrastructure market. As you beget heard from plenary the team members and that they increased the infrastructure revenue as total revenue percentage is significantly -- when you combine their change over the strategy and their team's tenacious execution, they expanded their grievous margin from low 53s to 63.5, 1000 basis point increase, it feels really benign to screech that because you don't accept many opportunities to screech you increased your grievous margin by 10 percentage point.
On operating expenses side, Matt talk about in detail how they approach results avocation. It's really particular data driven, result driven focused on returns, so when you spy at that and spy at the how the leverage of their model, they beget increased their operating margin from low single-digit to 30%. They are very proud about the model they beget viewed. This model generates a lot of a cash rush too. So when you spy at it, starting from Q2 fiscal '17 to Q2 fiscal '19 despite of their long respite over participate repurchase associated with the Cavium transaction. They beget returned a billion dollar cash back to shareholders, through participate repurchase and the dividend. They are very committed to revert cash to shareholders.
Let me switch gear to talk about their merger with the Cavium and their long-term monetary model. As their team highlighted earlier, this merger really increased their market opening from $8 billion to $18 billion. The infrastructure revenue as a percentage of total revenue increased to two-third. And also, they increased their skill and the diversification with the 200 million synergies, when you add plenary of them together, it truly accelerates their top line revenue grievous earning expansion and to generate higher intrinsic value for shareholders in the long-term.
So I'll select you through their revenue profile and assumptions under their long-term monetary model. Matt talked about earlier, if you spy at their core networking and storage market, they await the market SAM without a server SoC to grow 7% going forward. They are very well-positioned to address the market opening to grow Marvell overall top line revenue by 6% to 8%.
We await their storage industry to grow low single-digit largely in line with the market and then within storage, Dan talk about their opportunities and that they actually hold a very cautious assumption by the HDD market, which they await the overall market dollar SAM them to decline 7% but they await they continue to expand into the datacenter near land market which is growing double-digit they besides perceive their preamp revenue to continue to ramp. Those are their modern product cycles for Marvell. They will create incremental revenue opportunities to offset the market decline, so they believe their HDD industry is going to live flattish or decline single-digit going forward. Fiber channel market has been really stable and healthy, so they continue to await their fiber Channel industry to live flattish going forward.
Now on the glisten solution business, it's very exciting for us. Dan talked about the continued expansion to cloud enterprise datacenter. This industry design cycle actually it's very long, if you can recall Dan's charter to spy at the percentage of revenue they are going to enlarge in non-PC segment, those are the design wins they already secured and their team are really working on execution of those design wins, so they accomplish perceive revenue continue to grow driven by those opportunities. More importantly, they are very excited about the design wins their teams are working on to address great opportunities in both emerging embedded solutions and the glisten solution market. So, overall, they continue to await their glisten solution industry to grow inline or faster than market in the long-term.
Now let's switch into networking, the most exciting ingredient of their growth story. They accomplish await their networking industry to grow faster than market at the low teens going forward. Both Raghib and the Tom talked about the exciting opportunities they beget ahead of us. At the highest level, if you spy at both their processor product line and the internet product line, they are going to grow faster than market and their Wi-Fi product line largely is going to live in line with the market.
So from the end-market perspective, they perceive growth opening across plenary their End market, which are so exciting, and that there are so many different drivers. I'm going to only highlight a few key drivers. First is the enterprise, the upgrade cycle Raghib has talked about and if you recall Marvell's enterprise switch and the five businesses has been growing double-digit during the last few quarters and the Cavium side their enterprise industry has been growing double-digit too. When you combine the both portfolios, the leverage they beget and the design wins their team already won, they accomplish reflect their enterprise industry will continue to grow with a tenacious momentum.
Next on 5G, I'm pretty certain you guys plenary recollect Raghib's 4X chart that's what I remember, that's how my brain functioned. So 5G is really largest growth opening for Marvell going forward and then their current model assumption actually is just their lead customer maintain their existing market participate and their team are working really hard to deliver on the schedule of the design wins.
Next is datacenter, they beget a very unique set of very innovative product lines in datacenter ranging from a security to Ethernet offload to gearbox re-timer. So plenary of them when you spy at the in the overall context of Marvell today, which you know, it's over $3.4 billion revenue, the revenue base is miniature but plenary of them actually are going to beget a considerable opening going forward. The growth rate of plenary those product lines actually it's going to live higher than carrier space and enterprise space. So they accomplish await their datacenter revenue to grow most significantly and enlarge their presence in datacenter in the long-term.
Next their remark on the server processor business, they really only included very qualify revenue in their baseline model just based on current customer design wins and the customer engagement. I'll talk about later what they did not include. so other product line for modeling purpose, they continue to await other product line to decline high-single-digit.
Now let me talk about what are the upsides they can perceive going forward. First, revenue synergy; Tom talked a lot about the portfolio, the engagement they beget with the customers. They beget not included any revenue synergy in their baseline model. So if they generate a revenue synergy that will live tremendous upside. Of course, their industry has long cycle, so you should await it for the future business. Secondly, 5G in their baseline model, they only included the current design wins their team are working very hard to deliver, they did not comprise any additional customer design win especially on the baseband side. If they win modern designs there's going to live tremendous upside, I really exigency your sunglass in that time. That's very exciting. You don't accept CFO excited most of the times.
And so, the next one is Hyperscale ARM server adoption, so in their baseline model, as I said, they only comprise the design wins with the OEMs, you know some of the testing chips, any major hyper secured datacenter adoption is not in their current model, so there will live upside the too. So as you can see, they beget a lot of the exciting opportunities ahead of Marvell and their team is working really hard to execute because plenary of their businesses are very long cycle business, so everything I talk about here their team is executing forward.
Now let's switch gear to talk about synergy and besides talk about their operating expense model. So they reached their synergy to $200 million during their last earnings call, then they got the questions about why it's so high, why it's so high. So you guys never had, so now let me intricate to you why they achieved, they are going to achieve $200 million synergies.
So this is about 8% of the total company's disburse which comprise both the cost of sales and operating expense, which is actually within the scope of previous transactions but it certainly powers the high-end. The reasons they could achieve with these kinds of unique synergies because they accomplish beget unique opportunities. I will highlight it to you, first, it's on the R&D side. So both companies were investing in high-end the datacenter switch and embedded the server processor -- embedded the processor industry and a lot of you know, those are very great investment. So their team worked extraordinarily hard to consolidate the roadmap. So in each business, they pick the most competitive roadmap going forward which they End up saving a lot of money and contributed significantly to the savings over $200 million synergy.
The second one is facility consolidation. It's relatively smaller, but it's besides very significant for us. They beget a lot of overlapping offices across different locations, they can consolidate, but the most unique one is they beget their great Design headquarters and headquarter located nearby to each other and one of us betide to beget actual space. So they actually they are going to live able to plod the all Cavium headquarter and the great design team to Marvell's campus without enlarge in any space that consolidation is very efficient. So they beget been able to deliver more facility cost which is a fixed cost as you guys know to achieve synergy and contribute it to this $200 million revenue.
Now let's spy at their synergy execution timeline. Their team actually is executing ahead of time. On grievous margin side, they expected to achieve $50 million synergy which was started in Q1 fiscal '20 and over time to ramp up to achieve the total $50 million synergy. As they spy at their grievous margin, they usher their Q3 as combined the company grievous margin to live around the 64.5%. So if they achieve their $50 million synergy target, they will live able to accept it to around 66% grievous margin.
On the operating expense side, their combined company's operating expenses elude rate base is about $325 million and then they guided their operating expense for Q3 fiscal '19 to live in the scope of $300 million and $305 million. So what had implied is we're going to achieve $90 million elude rate of synergy out of the gate as a combined company. Then you are going to perceive some payroll tax and the merit enlarge in the first-half of fiscal '20, but once they migrated to second-half of fiscal '20, they are going to live able to complete one ERP program and achieve the remaining $60 million synergy for the overall $150 million synergy.
So when you reflect about how they manage their operating expense, Matt talked about extensively the discipline approach they have. On SG&A side, Marvell standalone already achieved below 8% SG&A as a percentage of revenue. They will continue to drive operational excellence to target SG&A as a percentage of revenue to live between 6% to 7%.
On the R&D investment side, they want to invest for the future, so they await their R&D investment ranges around 24% to 25%. The course they approach it as Matt discussed is for the established business, they will manage R&D expense to live much lower than 24% to 25% to maximize long-term cash flow.
For grievous business, they are going to invest higher than 24% to 25% to drive their industry grievous and the expansion. For strategic investments, those are the investments that we're creating future and to innovate, we're going to monitor those investments and focus on revert on those investments.
Automotive is a considerable example, in the past last year they did not comprise automotive into their SAM but the investment has been paying off and we're very excited about their opportunities in the automotive Ethernet market.
So when you effect plenary together to spy at their long-term monetary model, they await to deliver top line revenue growth in their baseline model to live at 6% and 8%, they await their grievous margin to live greater than 66%. On the operating expense side, they will execute certain they enlarge OpEx less than their revenue increases to accept to the leverage model. So operating margin, we're expecting to live greater than 35%.
Our free cash flow, the company has a very high free cash rush conversion because their CapEx is only 2% to 3% of revenue and the depreciation it's probably 3% to 4%. So it's very efficient that cash rush conversion model, they await their free cash rush to live approximately 100% of non-GAAP net income which is about 30%. So when you spy at this model we're building which is focus on infrastructure market, that's why they can beget this kind of a fundamental economics behind their model.
Now let me talk about the capital structure and the capital returns, so the course they reflect about the capital structure is to really beget a tenacious monetary flexibility, so they can invest for the longer term, that's how they structured a Cavium transaction.
So at End of Q2 fiscal 2019 after they closed the transaction, they beget over $500 million cash and besides they beget $500 million undrawn credit facility, their leverage is very reasonable, their grievous leverage ratio is about two times and the net leverage ratio is only 1.5 times.
As a company, it's very considerable for us to maintain their investment grade because they really want to invest for the future, so they want to beget ample access to that market at very reasonable cost, invest through economic cycles and besides select odds for consolidation opportunities if they present themselves. They are going to start to pay down their debt, since they generate a lot of the cash flow, they reflect of between their debt reduction draw and their EBITDA expansion they will live able to achieve 1.5 times grievous leverage ratio in the next 18 months in the next 12 months.
Now let's talk about capital returns. You beget heard from their team, they beget tremendous opportunities ahead of us. So their number one objective is to invest in their industry organically and through acquisitions, then their long-term objective is revert cash to shareholders is to revert at least 50% of free cash rush to shareholders through both dividend and the participate repurchase.
In the near-term, let me talk about their priorities, first their industry is generating significant cash rush even privilege now, so they accomplish reflect they can maintain their dividend flat and started to pay down the debt, the flat for debt reduction is going to live modified with their participate repurchase plan.
As of you'll probably beget seen is this money, their board of directors raised their participate repurchase draw to $1 billion by authorizing 700 million participate in repurchase, so when you spy at their numbers some of you may track it closely not is they actually already started to buying their shares after Q2 earnings call, they already bought back about $50 million of their shares, we're really buying their participate at this current flat as considerable investment.
If you spy at their long-term target model, if they can achieve their long-term target model, they will perceive huge upside. So in summary, when you reflect about the Marvell, they are very well positioned to address very great opportunities in the infrastructure market space and that they beget a very powerful industry model with underlying tenacious fundamental economics to drive long-term cash flow. And they besides beget a tenacious monetary position and a tenacious monetary flexibility, so they can both continue to invest for the long-term and then revert cash to shareholders.
Before I invite their team to approach up for mp;A, I just want to select this opening on behalf of their all team to really thank everyone for your interest in Marvell and besides thank their long-term shareholders for your tenacious support through their journey to build really considerable company. Their objective is to build the industry for the longer term and that they want to create a shareholder value for the longer term.
Thank you and I will invite their team to approach up for mp;A.
A - Ashish Saran
All right, folks. We're going to Go to the mp;A session. There are folks who beget wow. It's going to live interesting. Can they start on here, please?
He has got the mic.
Okay, Go ahead. Well, that was fast. Okay.
My apologies. Thanks for taking my question. I had a yoke of questions about the -- some of the products of Cavium that were rationalized or decided to -- you started to stick with. So first with XPliant, what inside of XPliant from an IP perspective helps to preserve product line maybe in the core enterprise End market, and then as well affecting into the datacenter End market? And with respect to ThunderX2, what was the thought process behind, or the considerations in deciding to maintain that? Was it because you basically had sunk plenary the R&D cost already, you decided to roll the dice, or is it indicative of the product having some benign long-term growth prospects?
Okay. So when it comes to XPliant product line, it is combining. They beget combined a roadmap of the two product line that they were working at Marvell, Falcon product line as well as XPliant, they were both targeted for the identical market. There are few areas -- blocks in XPliant, which is very interesting, for example, a supple capability of supple parcels [ph] and the all aspects of visibility and programmability, so being able to really beget the capability to view the packets through the network and so on. So they beget taken those thing and they beget merged the two architecture in the subsequent architecture. So that was the thought process behind XPliant.
I'll give a quick remark on Thunder. I live of value I beget never been a believer that you continue with the project, because of the sunk cost or looking at how much you've spent today to justify let's maintain going forward I wasn't raised that way, trained that way, nor believe that. So they very much spy at that opening as on the go-forward, and what's the market opportunity, how is the traction going, what's the customer adoption and plenary that's going very well as you heard from Raghib. So that one we're enthusiastic about.
The overall traction is really excellent, and we're making considerable progress, and that's how they umpire plenary of their businesses.
Thanks. privilege here, Ross Seymore from Deutsche Bank. Two questions; first one for Matt, during the last few months there was a lot of consternation about the Cavium revenue level. I know you went into considerable deal to detail in your last conference muster about that, but any updates as to either the claim side or the channel rationalization side that they can spy at going forward?
Sure. So yes, so they did their muster maybe a month ago. I reflect they gave the outlook. They gave -- they leaned in a cramped bit more than they normally do, right, to give investors visibility. And so, when they announce their next earnings, obviously we'll give you guys a bit on the short-term. Today's goal was really to focus on, "Hey, what are the long-term prospects of the company? Where are they taking it? And where are they heading?" So, not going to remark about the short-term today.
I'll Go with the long-term follow-up with Raghib…
- the 5G side, obviously that's an district you're very excited about that 4X multiple as they Go from 4G to 5G, talk a cramped bit about the mile markers and the timing of when they can start to perceive some of that evidence within your networking business?
So, with their already design wins and the lead customer, they are expecting revenue started later participate of the next year. And of course, it will ramp into calendar year '20. Now, in addition to that, as they mentioned earlier and Jean besides pointed out, they are actually heavily engaged with the lot of other OEMs. So they are actually in a very benign position to win additional design. So, their current base draw does not comprise any of the additional design, or does not comprise actually additional market participate gain by their lead customers.
Thanks, Blayne Curtis of Barclays. Two questions for Raghib or Matt, this is the first time you've mentioned that AI inference product, if you can give any notion on timing of that when you execute sample, and then obviously as everybody knows there's many public and private companies chasing after this market, if you maybe just talk about the differentiation, you beget been working on this, I reflect you said two years, why you decided as Mike commented, to Go forward with the product? And then, just following up on Ross's question on 5G, if you did win on these next-gen 5G base stations, just may live some timing as to when that could contribute? Thanks.
So, the product that they are going to sample, they are -- it's going to sample next year, right, in terms of timing of the product. Now, if you spy at this all market, as I mentioned earlier, there is a all kind of goal, right, sort of -- I reflect everybody is trying to really chase the tremendous guy out there, which is really focused on training. That is why everybody is trying to build the biggest and baddest sort of a thing, right?
Our approach is their working with the customer is that they really exigency a solution which is really designed for a scale, right, and that is why they -- when -- just like always that they did in Cavium as well, they create not only that they reflect that way, but other tremendous customers besides reflect that way, and that's how they started working with a nearby partner, and that's why we're developing the part.
Maybe I'll just add, I reflect we're well-positioned there as well, because in the End when plenary these chips are available and plenary the benchmarks are done, a huge consideration I reflect of any infrastructure company that's going to deploy AI processing into their hardware, into their systems, is going to want the identical dynamics that they described today, you know, great supplier, viable, focused on the End market, track record of execution. And so, I reflect that's going to End up playing to their strengths, but as Raghib said, it's an lively project, we're developing the chip, we're sampling it next year, we've done the evaluation of its prospects, right, relative to other investments we're making, we're very excited about it. And they reflect that although we're not including it in their SAM, because by the course the SAM at this point is actually -- these numbers are so great that it's hard to really -- you could accept third-party reports, so they prefer to execute the project, kind of drudgery with their customers, and as they execute progress cramped bit like they did with automotive Ethernet, then comprise this at a later date in terms of the SAM and the opportunity…
Blayne, and just to add a cramped bit more, actually in terms of plenary of the benchmarks, plenary of the metrics, and in terms of performance, power efficiency, cost efficiency, actually their solution what they know based on plenary the other available, is going to live the market-leading solution in the market.
There was a follow-up question I reflect on additional baseband customers, and maybe you can just remark timing -- you know, it takes a yoke of years.
Sure. Yes, I reflect that -- again, we're very focused on getting their lead customer production and enabling them to live successful. I reflect the ally model that they showed when they said plenary ASIC merchant partner, I reflect that's being very well-received especially in this 5G cycle, where there's a pretty great diversity of not only companies developing infrastructure for this, but the kinds of radios that they want to develop are besides going to live quite disparate. And so, just to accomplish a gigantic ASIC for each of these is going to live a huge lift, and I reflect the fact that the Cavium team actually has this proven track record on 4G and 5G IP, that ally model, but starts to spy very attractive. So anything they were -- they would live able to win there would clearly live a modern product development. They exigency to Go often start that chip. So these are not ones that would ramp any time next year, they would live in the future, but this again is a very benign long-term opening because as you can perceive from the 3G cycle, the 4G cycle, now the 5G cycle, these are multi-year. And if you talk to their customers, they reflect actually 5G is going to last even longer from a cycle point of view than 4G. So, these are longer term opportunities that would layer in.
Hi, Karl Ackerman from Cowen. Two questions please, I reflect the shift from media to flash-based storage is clearly a tailwind for you, but I reflect one of the concerns from investors is how quick SSD controller ASPs approach those of your hard drive controller? So, how should they reflect about sustainability of the premium you receive on glisten controllers versus hard drive controllers? Is it in a strictly linked to NAND ASPs per gigabyte?
And I guess as a follow-up, how they reflect about both the trajectory and potential competition between yourselves and the plenary glisten array providers as your Ethernet bunch of glisten seems like a considerable alternative for NVMe over fabric arrays? Thanks.
Okay, the first question was really ASP comparison between HDD and glisten controllers, was that the question?
How sustainable you reflect the premium of your NAND controllers will remain above hard drive controllers?
I see. Okay. So I reflect it depends on the segment that you're in. So clearly, with their focus more towards a datacenter and enterprise we're able to provide modern features, functions, and accept a higher value out of those. Certainly in the notebook or PC space, you'll perceive transitions as you change interfaces or -- but that's more difficult in that space, and that's why we're focused more towards the other end.
And as far as the EBOF, right, for instance the plenary glisten arrays, I would screech that those types of customers execute in the plenary glisten arrays and the enterprise customers are customers that we're focusing on with those products. So I reflect that's actually very complementary that they hope to perceive them adopt those architectures.
To add to what Dan said, right, recollect in the glisten controllers that they design typically takes three years to design into the customers, walkway the customers, tenacious IP in cloud datacenter. The ASP actually is a wider range, right, it's because you contribute IP to your customers. So overall, as I said earlier, when you spy at the more commodity character of pricing, that has nothing to accomplish with their controller pricing, which is very much embedded into the IPO provider for the customers.
John Pitzer with Credit Suisse. Just Dan a follow-up on the storage side of the market, can you talk a cramped bit about the wish of NAND glisten providers to actually accept into the controller market, and how you perceive sort of the merchant controller market, market participate developing over time? And then Matt, as a follow-on or second question, you haven't included Cavium synergy -- revenue synergies into the forecast, can you just talk about where they should perceive those synergies first? I know in the past, you've talked about, for example, Cavium using other people's PHYs and that sort of an easy switch, how accomplish they reflect about the build of revenue synergies over time especially given the long duration design cycles that you're involved in?
So I mean, obviously upright integration with the NAND vendors is that we've always seen, we've always considered in their SAMs that you know, they model based on some assumptions there as well. We've traditionally had industry with those customers. They continue to drudgery with them on certain industry models. So I reflect it's a natural participate of the evolution of that market. Clearly on the accomplish It Yourself models and some of the stuff we're doing at datacenters, you've got to recollect that they're going to focus a lot of their investments on some of the higher volume segments, but yet when you exigency some of the IPs and when you spy at going from screech 28 to 12 to seven nanometers doesn't that kind of expose that graph of how much expenses that gets to be. Marvell leverages across many, many chips. So there are certainly segments that build accomplish stuff themselves, there are segments where they will work, they will buy something directly from Marvell, or working some of these ally models and the industry models that -- again that showed. So I reflect it's something that they deem when they spy at their models, whether it live the SAM or the revenue.
Yes, just one quick remark on that, I reflect they beget been operating in that environment since I joined. I live of value they had -- I reflect the first month I was here, they had a management consulting company offered to accomplish a free month of drudgery for us, you know, to assist us accept integrated. And so, one of the studies they did was on their product lines, and they said, "Hey, you know, the SSDs product line, they recommend you shut it down because everybody is going vertical." And then they said, "Okay, well, let's really spy at the data," and of course that would beget been a horrible decision, because that became a hugely considerable industry for the company. So I reflect they are going to co-exist and I reflect as Dan mentioned, it's going to live different industry models.
On the revenue synergies, the tremendous opening there really is the design wins. I live of value there are -- there is not a lot of discretionary industry they beget in their portfolio, and I reflect that's actually a benign thing by nature that they are typically highly proprietary sole source. So, going out and chasing a customer through distribution or something is typically tough especially if you are trying to sell an Octeon. I reflect that's a challenge, right, but I accomplish reflect there are some tactical things that they are doing certainly in the short-term, but the overwhelming revenue synergy that they can achieve is actually from the examples that Tom gave, which is really going in as a combined company leveraging the combined IP of the two, and really selling a all system solution versus just selling a chip. And the pull-through is very strong. When you beget a processor, you know, a tenacious processor portfolio, it does enable you that sort of insight into the decision-making on the board at the earlier stages. So I reflect that's really where you should about it, and I reflect to the extent they are successful in doing that, then they will just push their revenue growth long-term at the higher End of their model.
Hi, Matt. Vivek Arya from Bank of America Merrill Lynch. Thanks for hosting the Analyst Day today. So first, near-term, I understand you don't want to talk about Marvell specifically, but there is a lot of worries about just the semi conductor cycle broadly, you know, China trade tension, CPU shortages, so to whatever extent you could, you know, assist effect their understanding at ease about how are you seeing just the broader psychodynamics play out? But longer term, give us rather of a report card about the integration process, because Cavium was running a very different way, right, it's very entrepreneurial company, but the pricing dynamics, lead times, customer engagements were done in a very different course than what they were used to seeing in Marvell, especially when after you and Jean took over. So, give us some sense of report card that are you through that integration process is now being elude as one smooth company, with one united front of customers, so that they can live assured that there is going to live the privilege execution behind some of the targets that you beget laid out? Thank you.
Sure, great, great. Two questions, so, yes, the first one I reflect is pretty interesting, right, to host an Analyst Day in the middle of -- about as clamorous an environment as they beget seen in a longtime, and I reflect plenary those issues are very actual and upon us in terms of the geopolitical issues and the global issues with respect to tariffs. It's unclear what ramifications and plenary that are going to be. They are silent working their course through that. I reflect on the PC one, I reflect to the extent that there is -- they are downstream from that, right, so I reflect the best source of information would live Go talk to the PC makers, or even the drive guys, right, and they are two levels removed. And so, they are going to just react, depending on what happens, but you certainly reflect any shortage there has to live short-term in nature just because at some point there will live enough PCs to supply the world, I reflect the world without enough PCs doesn't execute sense to me. So I reflect they will drudgery through that issue, okay? And they will deal with it.
I reflect on the integration, I will execute a few comments and then maybe I will let Raghib remark too since he came from the Cavium management team, he assist build the company, but they are very far long in integration. I reflect the fact that -- I live of value I showed the picture for a reason, they were very purposeful and thoughtful about the fact that it was a different culture and a different kind of company, and I reflect they took steps early on to obviously give the Cavium leaders significant roles in the combined company, they took significant steps to enroll them in their culture and walk them through kind of how they elude their business.
Our point of views has been very well-received. I reflect -- and they beget really -- I reflect integrated them well into their system, I live of value the fact that they did this portfolio review, which is really one of the most considerable processes they beget in their company. They disburse a all week doing that, and they probably had -- I don't know, eight or 10 Vice Presidents -- technical by the way, most of them are either follows or engineering leaders from Cavium actually link and assist us review the all combined set of businesses. And then at the End they made decisions about how they are going to earmark their resources, you know, these guys are smart, right, they accept that. "Okay, wait a minute. This industry is really not making as much money as they thought," or this one looks like it doesn't beget enough resources, or you know, and so when you actually select the tenacious technical acumen of Cavium and you combine it with the tenacious monetary acumen on Marvell, they view that as a very intertwined capability in their company. I reflect it's been very well-received. And so, the last point in that would live Tom has brought in the sales organization, they are pivoting from kind of commission-based, design win based, right, from an operations point of view, Andy Micallef, who is their leader there, he has done a considerable job in driving Marvell's operational excellence, right, that all team now reports with him. And I reflect people perceive the benefit to running it this way. So, maybe Raghib will give you a remark since you are…
Yes. So I want to screech overall integration is going on very well, and let me expose you although at outside it looks very different company, but when it comes to core engineering DNA, what company has similar DNA? They are hard working and engineers focused to create in a course to create value out there, right? So that is a benign thing.
The other aspect which actually worked out really well in this situation, the course top-down the Marvell management actually provided the environment, right? So it was very inclusive and very open. It's not like, okay, it's my way, what's your course sort of a thing. It's just like, okay, let's accept together, pattern out what is the privilege thing to do, which one, as Matt mentioned, which are the -- when you are merging the businesses, which makes sense to keep, which makes sense to let Go and things like that. So, plenary those decisions were very inclusive, and the team really feel like, "You know, I did participate of the overall thing." So, not only at the top level, but besides at various engineering flat and so on. So, overall environment is very, very kind of cooperative. They perceive the value of overall objective-driven and goal-driven method in which actually, is really appreciated by a lot of folks in the Cavium side as well. They accomplish realize that the scene are not done based on politics, but based on overall bigger company goals. So, those values actually are taken really well.
At the identical time, it is not that -- you know, entrepreneurial thing is kind of mash or Kill at all. I live of value if you really spy at character of products they talked about today, I mean, they are investing in a lot of considerable potential innovative products, right? So, I would screech the overall thought-process of being innovation-driven and being leader in certain areas is besides appreciated on overall management side. That is working out very well.
I mean, select one last question before they exigency to wrap up.
Quinn Bolton with Needham. Just a follow-up on 5G, you beget talked about the 4x enlarge as you Go from 4G to 5G, so I reflect your content opening is probably approaching a $1,000 per base station, and you beget talked about embedded processors or layer four to seven, you beget got the baseband processors, you accept your Ethernet switches and PHYs, where does the value approach from? Is it fairly evenly spread across those different solutions? Does the baseband processor portray the largest percentage of the dollar opening for base station?
Okay. Should I select that? So first of all, I thought you guys in finance can accomplish a better math, but anyways, just to repeat, in 3G they worked up for a $100 content, in 4G they increased it three to four times, and it's going from 4G to 5G, they are increasing another 4x compared to 4G, right? So you can accomplish the math.
Now, on the other -- on the side of where it is coming from, in reality it's a mix, it is more -- of course the processor side, the baseband processor as well as the Octeon [ph] processor is higher ASP, but then the additional -- you know, significant addition with the combination of switches and PHYs, because now this time the bandwidth which is affecting data you know, ply data by the base station, bandwidth is much higher, and as a result of that, you exigency switches which is at a much higher bandwidth flat compared to the before previously.
All right, well, thanks everyone for attending the session today. They really esteem the time you spent with us, and for folks in the room, there is a lunch next door. You will beget the haphazard to network a cramped bit more with the executive team there. So, thanks again.