Ladies and gentlemen, thank you for standing by, and welcome to Marvell's First Quarter Fiscal Year 2021 Earnings Call. At this time, all participants are in a listen only mode. After the speaker presentation, there will be a question and answer Please be advised that today's conference is being recorded. I would now like to hand the conference over to your Vice President of Investor Relations, Ashish Sharan. Sir,
Thank you, and good afternoon, everyone. Welcome to Marvell's First Quarter Fiscal Year 2021 earnings call. Joining me today are Matt Murphy, Marvell's President and CEO and Jean Hu, our CFO. I would like to remind everyone that certain comments today may include forward looking statements, which are subject to significant risks and uncertainties that could cause our actual results to differ materially from management's current expectations. Please review the cautionary statements and risk factors contained in our earnings press release, which we filed with the SEC today and posted on our website, as well as our most recent 10 ks and 10 Q filings, we do not intend to update our forward looking statements.
During our call today, we will refer to certain non GAAP financial measures. A reconciliation between our GAAP and non GAAP financial measures is available on our Web site in the Investor Relations section. With that, I'll turn the call over to Matt for his comments on our performance. Matt?
Thanks, Ashish, and good afternoon, everyone. During the Q1 of fiscal 2021, we delivered strong financial results and achieved 6 $94,000,000 in revenue, dollars 14,000,000 above the midpoint of guidance. As you may recall, our revenue guidance for the Q1 included an assumption for a 5% headwind from COVID-nineteen. We did, in fact, experience some impact on revenue from COVID-nineteen related issues, but stronger than expected demand from our data center and 5 gs infrastructure end markets and our networking business drove consolidated company revenue above the midpoint. Compared to our assumptions when we provided our revenue guidance for the Q1, the impact of COVID-nineteen turned out to be greater than expected in our storage business and lower than expected in our networking business.
Our GAAP loss per share was $0.17 Our non GAAP earnings per share was $0.18 above the high end of our guidance range, driven by higher revenue and tight OpEx control. On a sequential basis, revenue declined by 3%. However, compared to 4th quarter results adjusted for the divestiture of WiFi, revenue for the ongoing business was actually flat on a sequential basis, a very solid result under current conditions and in what is typically a seasonally weak quarter. I'm extremely pleased with our team's ability to deliver great results despite the massive disruption from the COVID-nineteen pandemic. The vast majority of our employees have been working from home since early March, earlier than most official shelter in place orders were announced, as the health and welfare of our employees and their families was our first priority.
Our IT team quickly jumped into action to provide secure remote access for our 5,000 plus employees. Our worldwide operations group battled through a number of supply chain challenges and met the upside in demand in our networking business. Since the start of work from home, our engineering team has taped out 8 new chips, an extremely high level of engineering output in this time frame, on schedule to meet our customers' critical time lines. Our sales team maintained a strong focus on customer communication through this critical time period and continued to drive our growing design win funnel. I have personally experienced a much higher level of direct engagement with the leadership at multiple key customers at the CEO level.
Normally, I fly out and meet my counterparts in person a few times a year. But during this crisis, we have quickly adopted virtual meetings over video to increase the frequency of communication. We believe the change in cadence reflects our customers' desire to involve us deeply as a trusted partner in solving the need for higher bandwidth and storage in the post COVID-nineteen world. Our mission as an infrastructure semiconductor solution provider to move, store, process and secure the world's data has never been more relevant, and we are seeing strong demand drivers for our products and technology. I truly believe that the culture we have created at Marvell is enabling our success and has earned us the trust of our employees, customers and partners, and our new brand reflects these values.
Celebrating the company's 25 year anniversary, we recently launched Marvell's new brand identity, marking a milestone in our transformation to focus on semiconductor solutions for data infrastructure. This was something we had considered doing soon after we started to pivot the So instead, we focused first on creating our new strategy, revamping our culture, building operational excellence and delivering business results. After 4 years of hard work, we were ready to reveal a new brand that mirrors our culture, which focuses on the substance first. Our brand is built on a foundation of collaboration and partnership with our customers, and we are driven by the belief that doing things the right way matters. Essential technology done right reflects our customer centric approach to high quality tailored solutions.
None of this would be possible without our dedicated employees who have been putting in a lot of extra hours through this unusual period, which is likely to persist for a while. Recognizing this challenge, we are giving all employees a few extra days off, and we are shutting down for a 4 day recharge weekend in mid June. Similarly, we have taken steps to help our contractors and the local community. We are continuing to pay our hourly workers in full as well as the service providers that support our facilities regardless of whether they can fully perform their duties. In addition, Marvell has started the COVID-nineteen relief fund to support our global communities and provide financial relief to those in need.
A portion of the fund is dedicated to matching employee donations that employees will also have the opportunity to participate. Until we are able to return to the office in Santa Clara, we plan to donate our entire cafeteria grocery budget to the Second Harvest Food Bank in support of those in need. It is in times of crisis that our values and leadership principles are truly tested. I'm proud of our team and the company we have built. We are in the fortunate position of being able to take care of our people and give back to our communities when it really counts.
During and after this crisis, we expect the demand for bandwidth to continue to grow stronger. Major cloud and service providers are facing unprecedented demand for their services as a result of so many people working from home and are scrambling to add capacity to their networks. 5 gs has become a strategic priority for many nations and also an economic growth driver for economies worldwide. Enterprises are rapidly upgrading their networking infrastructure to support their workforce with remote access to critical business applications and advanced collaboration tools in a secure and scalable environment. We're in the midst of enabling all the key megatrends in 5 gs, data center cloud, enterprise and automotive as a leading supplier of data infrastructure semiconductors.
Over the last few quarters, we have discussed in detail our progress in the carrier market tied to the transition to 5 gs. Investor focus has been high in this area, and I'll provide color on this opportunity in a few moments. For now, I'd like to draw your attention to the cloud data center market, which represents an opportunity for our storage and networking products similar in magnitude to 5 gs. This is in addition to our ARM server opportunity, which we have previously outlined as a large multibillion dollar market for us. We've been working on building our business through organic investments over the past 4 years, as well as through products and engagements via the Cavium and Avera acquisitions.
We are now starting to see the benefits from these efforts, and I'm excited to share them with you. In our storage business, over the last few years, we have been transitioning our HDD business focus on nearline drives, which are primarily consumed by cloud data centers, and this business has been growing nicely for us through both our storage controllers and preamplifiers. We expect that the 16 terabyte transition this year will drive strong growth for us. Our custom flash controllers are very well suited for cloud customers who have unique storage applications and are looking to disaggregate their SSDs to optimize performance, and we pioneered the do it yourself model. We also have storage accelerators in evaluation at multiple cloud customers.
In virtualized environments and cloud data centers, multiple customers can be accessing the same physical storage resources simultaneously. To ensure quality of service, expensive CPU cycles are consumed in managing workloads to ensure that all clients get uninterrupted access to their stored data. Our NVMe storage accelerators enable a better solution by offloading storage IO and freeing up CPU cycles, provisioning bandwidth and ensuring low latency to satisfy service level agreements, which are difficult to achieve with software only solutions. We expect to see initial production ramps for our storage accelerators starting later this year. In our networking business, our OCTEON compute platform is ideally suited for a variety of offload and acceleration applications in hyperscale data centers.
At GTC 2020 last week, Jensen Huang referred to this category of computing as the DPU or data processing unit and indicated it would represent the 3rd major pillar of computing in hyperscale data centers after the CPU and GPU. We couldn't agree more. OCTEON is an ideal VPU platform as it is an SoC that combines a scalable multi core ARM processor with high performance networking and a variety of hardened acceleration engines. As many of you know, Cavium was a pioneer in this area, and we have continued investing in this segment since the acquisition. Our recently announced OCTEON TX2 was specifically designed for cloud applications, such as our Liquid IO programmable SmartNIC and Liquid Security HSM or hardware security module product lines.
Cloud infrastructure requires optimized hardware for performance, cost and power at scale. Data center operators can achieve these goals by moving specific workloads onto highly optimized offload engines, leaving the host server to run more efficiently. Our Liquid IO SmartNICs offload network virtualization, storage protocols, security, distributed firewall, software defined networks and a network overlay for various cloud data center models. Our latest LiquidIO SmartNIC incorporating OCTEON TX2 is ramping now as cloud provider demand is surging. Our liquid security HSMs, which have been adopted by multiple leading cloud providers to provide fully managed cloud hosted protection for cryptographic keys and crypto as a service are also ramping.
Last year, we announced deployments with Amazon, Google and Oracle, and I am pleased that we have recently added another large Tier 1 cloud customer. Our ARM based server processors have been qualified at Microsoft for internal Azure development workloads, and we are looking forward to broader adoption in production environments. Our ThunderX 3 Processor will become available later this year, and we continue to engage with additional cloud providers on our solutions. The Avera acquisition brings to us leading edge custom chip design capability, which continues to become more relevant as hyperscalers seek to differentiate and control their own key pieces of silicon. They have very specific needs, and we can combine our leading intellectual property with their own unique IP to most efficiently solve the most complex system design challenges.
As these customers continue to design their own silicon, they need a reliable ASIC partner that they can count on for the long haul. Avera was already shipping into the cloud market when we acquired them, and they brought over additional cloud design wins, which have yet to go into production. The level of engagement with hyperscalers has also continued to grow after we completed the acquisition, and we are seen as a very attractive choice for next generation design decisions. Collectively, we are now building a strong cloud franchise combining our storage, networking and processor technology and have the unique ability to deliver products as standard, custom and semi custom solutions. I'm pleased to report that for the first time, we estimate that our revenue from cloud market exceeded 10% of total company revenue, growing from a fairly small position at the same time last year.
We expect that this business will continue to grow with that end market and our own unique product cycles and become another key growth driver for Marvell. Let me now move on to discussing our 2 businesses in more detail. 1st, in our networking business, revenue during the quarter was 394,000,000 dollars and grew 5% sequentially, significantly above our expectations for a low to mid single digit decline. Compared to Q4 results adjusted for the divestiture of WiFi, growth for the continuing networking business in the Q1 was in the double digits sequentially, and I'm thrilled with this result. Excellent supply chain execution allowed us to meet strong demand from cloud data centers for our smart NICs and security adapters that I discussed earlier.
Our 1 gig and 10 gig PHY products experienced robust demand from enterprise data centers as companies continued to upgrade the core of their network operations to deal with higher security and throughput demands. Our wireless infrastructure business grew sequentially and performed better than our expectations of a flattish quarter, in strong contrast to pause or slowdown in 5 gs deployments mentioned by a number of industry peers for the same time period. We are benefiting from the start of the 5 gs build out in China, where we have content through 2 of our key customers, who collectively have been awarded about 50% of the share of recently announced contracts. We are currently supplying custom ASICs and standard connectivity products to these customers and are working with them on additional opportunities. In addition, we have a very strong position at Samsung, where we had just started to ship our full suite of solutions and expect that their success in Korea, Japan and the U.
S. To drive growth in the coming year. Our custom baseband solution for Nokia progressing very well through qualification, and we expect to start production later this year. Collectively, we are designed into the latest generation of base stations into multiple OEMs, which are expected to be powering all key geographies over several years. We are still at the very beginning of the industry transition from 4 gs to 5 gs and look forward to driving significant revenue growth from this end market.
Demand from the enterprise campus end market remained stable on a sequential basis, a good outcome under current market conditions. When we start to come out of the COVID-nineteen crisis, we are looking forward to the deployment of WiFi 6, which requires higher speed Ethernet connections. We expect this adoption will drive growth for our multi gig Ethernet products as we have a very strong design win position across most leading OEMs for enterprise Wi Fi access points, and we also have a strong position on campus switches, which aggregate traffic from these access points. We recently taped out additional products from our recently announced OCTEON TX2 family of ARM based infrastructure processors, including our flagship highest performance solution featuring 36 cores, which can deliver up to 200 gigabits per second of packet processing for the most demanding applications. We also taped out our fusion based processor customized for massive MIMO applications for our leading wireless customer ahead of schedule.
Our customer will be using this solution in their high volume massive MIMO SKUs, and we expect to be ramping into production in the fall. The design win funnel for OCTEON Processors is continuing to expand with a broad set of Tier 1 networking and high end security firewall OEMs. The OCTEON platform is an extensive ARM based embedded processor product family, which can scale from 4 cores at the low end to 36 cores on the high end, all on a common software framework. This is a critical advantage for Marvell when engaging with Tier 1 networking OEMs who want to partner with solution providers able to support their broad range of products, allowing massive software reuse. We are deep into the development already of our next generation OCTEON family based on advanced node process technology.
The combined strength of our recent OCTEON announcements and aggressive road map is resonating very well with key OEMs. In our emerging automotive business, there are encouraging signs that auto production is restarting in multiple geographies. And as this happens, expect to start ramping Ethernet switches in PHYs in the second half of our fiscal year for the designs we had won in model year 2021 vehicles. We expect to grow our automotive footprint over the next few years through our Ethernet connectivity products, but we see an even bigger opportunity for Marvell as cars continue to evolve towards becoming truly connected platforms. Traditional point to point solutions in cars are heading towards zonal architectures, very similar to modern enterprise networking, where endpoints connect to centralized compute and storage resources over a secure Ethernet backbone.
We believe that we have all the pieces in place to be a leader in the market for this evolution: low power and high performance ARM based embedded processors, proven security IP, automotive grade gigabit and multi gigabit Ethernet connectivity, advanced storage technology and a solution delivery model, which can span from standard products to full custom. We will discuss this broader initiative in more detail at our Investor Day in October. Let me now discuss the outlook for our networking business. After a very strong Q1, we expect our networking business to continue to grow in the Q2 of fiscal 2021 and project revenue to increase in the low single digits sequentially. We expect this growth to be driven by 5 gs wireless infrastructure deployments in our carrier end market.
Looking out to the Q3, based on our recent bookings and firm backlog position, we expect our 5 gs business to continue to ramp strongly. Turning now to our storage business. Storage revenue for the Q1 was 259,000,000 dollars a double digit sequential decline, worse than our expectations for a mid single digit decline. This was due to multiple quarter progressed, impacting production of fiber channel adapters and hard disk drives. Our SSD controller business, in contrast, remained resilient and grew sequentially, driven by higher shipments into data center applications.
Looking to the 2nd quarter, we expect our storage business to start to recover from COVID-nineteen impacts on the supply chain, although a full recovery is more likely in our Q3. We have started initial shipments of our custom SSD controllers for system level applications in our 2nd fiscal quarter. As a result, we are projecting storage revenue growth of approximately 10% sequentially for the 2nd quarter. This custom SSD controller is based on our new PCIe Gen 4 architecture and was developed in close collaboration with our customer to create a storage controller uniquely tailored to their application. I'm very pleased with our strong engineering and operations execution, which has enabled us to start early shipments to meet our customers' production schedule.
We expect the storage controller to ramp into high volume in our 3rd fiscal quarter. In closing, we have a strategy focused on semiconductor solutions for data infrastructure. The critical nature of that infrastructure became even more evident during the current crisis as it keeps people connected, businesses running and information flowing. Our pivot to data infrastructure wasn't an accident and is a direct result of our decisive strategic portfolio actions over the last 4 years. Our strategy is in place, we are executing well, and we have the right team and technology to exceed our customers' expectations.
Our booking and backlog trends are very healthy, especially in 5 gs, cloud and custom SSD, and support our expectations of driving overall company growth beyond the Q2. We, of course, will continue to closely monitor demand signals from our customers in this uncertain environment. We remain confident in our ability to continue to benefit over the long term from the macro trends I outlined during this call. With that, I'll turn the call over to Jean for more detail and outlook.
Thanks, Matt, and good afternoon, everyone. I'll start with a review of our financial results for the Q1 and then provide our current outlook for the Q2 of fiscal 2021. Revenue in the Q4 was 6 $94,000,000 above the middle point of our guidance. Networking represented 57% of our revenue in the Q1 with storage contributing 37%. Revenue from other accounted for 6% of our revenue.
As a reminder, this business consists of products such as printer solutions and application processors that we have stopped investing. So we expect they will continue to decline over time. Our guidance for the Q2 anticipated an approximately 20% sequential decline in revenue for these products. GAAP gross margin was 47.1%, which includes the amortization of both Aquantia and Avaira inventory step up cost. Non GAAP gross margin was 62.8 percent of revenue and reflects the change in product mix compared to our expectations as Matt discussed earlier.
GAAP operating expenses were 423,000,000 dollars Non GAAP operating expenses were $300,000,000 lower than expected, primarily because of our continued focus on OpEx management and the lower travel expenses due to the COVID-nineteen impact. I'm pleased that we were able to get to our target for non GAAP operating expenses 3 quarters earlier than expected. And we currently intend to manage OpEx at this level through the rest of the fiscal year. GAAP operating loss was RMB96 1,000,000. Dollars Non GAAP operating profit was $136,000,000 or 19.6 percent of revenue.
For the Q1, GAAP loss per diluted share was $0.17 Non GAAP income per diluted share was $0.18 about the high end of our guidance range. Now turning to our balance sheet. During the quarter, cash flow from operations was $176,000,000 We returned $65,000,000 to shareholders through 25,000,000 dollars in share repurchases and $40,000,000 in dividends. We temporarily suspended our share repurchase program midway through the Q1 as we believe it's prudent to further strengthen our liquidity and increase our cash balance during the uncertain environment. We'll continue to evaluate the business conditions to decide when to restart the re share repurchase program.
We exited the quarter with $660,000,000 in cash and short term investment And we have $500,000,000 liquidity available from our undrawn revolver. Our net debt to EBITDA ratio was 1.2 times on a trailing 12 months basis. I'm very pleased with our solid financial performance in the Q1 and the current environment. We achieved this result by building a strong business model to focus on the growing opportunities in the infrastructure market and continuing to be very disciplined in operating our business to focus on return on investment, while keeping ample liquidity and the low debt leverage ratio. Our performance also underpins our strong investment grade credit profile, which provides us with a significant financial flexibility to invest in our long term growth initiatives, so we can emerge from this uncertain time even stronger.
Now moving on to our current outlook for the Q2 of fiscal 2021. We are forecasting revenue to be in the range of $720,000,000 plus or minus 5%. We anticipate our GAAP gross margin will be approximately 50.6% and our non GAAP gross margin will be approximately 63%. We project our GAAP operating expenses to be approximately 393,000,000 dollars We anticipate our non GAAP operating expenses to be approximately 300,000,000 approximately $15,000,000 and we expect non GAAP tax rate of 5%. As a result, we anticipate GAAP loss per diluted share in the range of $0.02 to $0.10 and a non GAAP income per diluted share in the range of $0.17 to $0.23 Operator, please open the line and announce Q and A instructions.
Thank you.
Our first question comes from the line of Timothy Arcuri of UBS. Your question please.
Thanks a lot. In the release, you had mentioned that you had some impact for the trade, for what's happening in trade. So Jean, can you just talk about that and talk about maybe what the impact is and maybe whether there is a longer term opportunity, maybe now that high silicon is going to be quite restricted in sort of what they can build, will there be an opportunity for you there? And I guess also, I just wanted to ask whether you have an update on the 6 $1,000,000 incremental 5 gs opportunity. It sounds like it's going to be quite a bit bigger than that, maybe as big as $1,000,000,000 So I'm wondering if you can comment on that.
Thanks.
Yes. Hi, Kim. Thank you. Yes, on the trade impact, I think the impact from Huawei, it has happened in the past. So we it's already impacted our revenue last year.
We actually almost have a non Huawei revenue anymore going forward. On the other customers right now, based on our assessment, we don't see additional impact. And I think the way to think about it is there it's uncertain, but right now we don't see other impact. I'll let Matt answer the 5 gs question.
Sure. Yes. And Tim, just to clarify on the trade side. Yes, Gene mentioned, we felt primarily the brunt of that impact about a year ago when the initial restrictions were put in place. And so we've worked through that issue.
I think the other thing you may be asking is what's the kind of overall picture if there's an impact to that particular customer in the 5 gs space and will other vendors share? Will that be a tailwind for Marvell? And the answer to that is in our current base case assumption, we're assuming no share shifts. We have strong position at the other 4. But certainly to the extent that there are share movements within the 5 gs players that would be to the benefit of the customers that we supply.
That would be a good thing for us. Related to that was your second part of the question, which was on the long term opportunity. Yes, it's actually a little bit larger. It was $600,000,000 when we it's standalone Marvell and then when we did the acquisition of Avera, we added another 1 $150,000,000 of run rate business coming over. So that really framed the whole opportunity at $750,000,000 We still feel very good about that.
As you heard in my prepared remarks, as well as even some of the comments from last quarter, we continue to have very strong design win momentum there and content increase at a number of the key players. And we continue to be heavily engaged and we're still in the early stages of this and we hope to update that figure at some point in the future. But just clearly based on what we've won already, there's 750 that's in the pipeline, probably a little bit more if you look at the Nokia partnership that we announced last quarter. So that's very much on track. Thanks for the question by the way.
Okay, awesome. Thanks.
Thank you. Our next question comes from the line of Gary Mobley of Wells Fargo Securities. Your line is open.
Hey, everyone. Good afternoon. Congrats on a good start to the fiscal year. I want to start out by asking about the strength that you've been seeing on the networking side, in particular, 5 gs and cloud as well. And so clearly, we've seen some good demand trends there.
And I'm just curious to hear your perspective on where we might sit with respect to some of those customers, maybe adding some capacity ahead of demand or are they playing catch up? And just trying to get
a sense of whether
or not there might be some carry through on the strength into the second half of the year.
Sure. Yes, Gary, this is Matt. I'll take that. So yes, we're off to a great start in our Q1. As I indicated, last quarter, there was we had a lot of momentum in our business exiting the year.
I would also give you a higher level framework to start with, which is remember that the overall semiconductor industry in 2019 went through the 3rd largest down cycle we've seen in the last 25 years, primarily on the back of the trade tensions. And so coming into calendar 2020, just kind of broadly speaking, inventories were low in the customer base and supply chain. Certainly, even at Marvell, our own inventories through our channel and our customers were low. And so the demand that we had seen exiting the year even before COVID-nineteen hit has continued. And the drivers we see in our business are we think very sustainable certainly in the two areas that you mentioned.
We're at the early stages of 5 gs and you've got the China market that is expecting to have very robust demand this year. That's all very sort of publicly announced and we're at the beginning stages of that. We expect other regions to continue to follow, which would be very good for our customers. On the cloud side, I think that demand trend is certainly here to stay. And while there has been some talk about potential digestion and things of that nature, at least from our lenses, Marvell, we have very unique product cycles that we mentioned both on our SmartNICs, our brand new Liquid iO3 SmartNIC product, which is ramping very strongly, our liquid security products, storage, I think there's a strong view on the nearline side that that will be a very good year.
So overall, at least from Marvell, we think those are very sustainable. Last comment I would make is in Q1, we had the additional benefit of some very good enterprise strength, primarily in enterprise data center and that helped us as well. So I think we've got a nice setup of various moving pieces, but certainly from a sustainability standpoint, we see 5 gs and cloud really driving our business this year through the remainder of the year.
Okay. As a follow-up question, I want to ask about China and 5 gs. You mentioned some success in market share gains for some of your customers in that region. Based on their tender activity and whatnot, when would you anticipate a more meaningful contribution from that specific region?
Well, it's in a little bit started and helped us even in the Q1 as that began. We'll see more in Q2 with really we see a very, very strong Q3, our Q3 for that particular business. And I would say that's really when the volume starts. So we're still in the early stages of that build with some of that comprehended in our Q2, but a much stronger outlook for Q3 and beyond. Got you.
I appreciate it. Thanks.
Thank you. Our next question comes from the line of C. J. Muse of Evercore. Your line is open.
Yes, good afternoon. Thank you for taking the question. I guess another question on China. Could you walk through your infrastructure exposure there both directed through Avera and potentially additional opportunities outside of Huawei? Thank you.
Sure. Hey, C. J. I'll start with that and then I'll ask Jean to add if she'd like to. So a couple of things to note.
We definitely have indigenous China revenue. We prior to the entity list issues, Huawei was our number one customer in China. And I think that's been very well telegraphed what's going on there. As you heard my commentary, we've ended up gaining some share at their biggest competitor in China. And so that's going to serve us really well in the 5 gs ramp.
And then additionally, we supply a number of other OEMs. We haven't broken that number out. But what I'd say is, if you look at our historical Qs, they just have shipped to numbers, like 50%, 40% some percent. And that's obviously way overstated because that reflects primarily contract manufacturing activity in the region. So we've got, I think exposure in the right places there, both in 5 gs as well as some of the key enterprise OEMs there.
And then we have some distribution as well. So I think our business is reasonably diversified outside of the first customer I mentioned.
Thanks, P. J. Can we have the next question, please?
Our next question comes from the line of Vivek Arya of Bank of America. Your line is open.
Thanks for taking my question and congratulations on the improving growth and the execution. Matt, just a clarification, I think you started to mention something about networking continuing growth in Q3. I was wondering if you could elaborate on that. And then the question really is, have you seen any double ordering or excess inventory or any abnormal orders from any of your customers in cloud or data center or the China markets? Because I think we conceptually get that certain parts of your customer base are strong.
But just given all the macro issues, I think investors are naturally worried about anything abnormal that might be going on because of all the trade concerns that are ongoing. So any color in terms of actual consumption of your products, I think, would be very helpful. Thank you.
Yes, great. Let me start with Q3 and then I'll move to the over ordering question as well. So on Q3, just to be clear, at the highest level, what I said in my prepared remarks was that we see we do see overall company revenue increasing from Q2 to Q3 at this point given the visibility we have. That's overall company. And that's driven by primarily our networking business, which is 5 gs and cloud, but also by this custom SSD opportunity I mentioned.
And so we see strength actually beyond Q2 both in our networking business and in our storage business. On the subject of double ordering and the concerns out there, look, I've been on this rodeo for a while. I've seen various shocks to the system and then the resultant supply demand imbalances. And I'm very acutely sensitive to those and watching those like a hawk. From our point of view, I think the interesting dynamic on this one, Vivek, as I mentioned earlier, is that the industry exiting 2019 was coming out of a very severe downturn where customer inventory levels were low.
And so certainly there's been a they probably overdid it. And I think even certainly without COVID-nineteen, I think we were going to see even probably stronger demand as we indicated for last quarter. So that's one part of the setup. And the second is, we've at least in Marvell, we've really over indexed on this in terms of spending time on this matter and trying to look at all the customer ordering patterns, does their market share add up to more than 100%, how much do we think is buffer versus not. And the end conclusion is I'm very comfortable with the current forecast we have and also backlog and order position that we've got.
And I'd just make another note, which is as well, just to hedge it a little bit, I mean, we've been running our distribution inventory well below kind of our target levels. And part of that is just to ensure that we maintain integrity in the demand side and make sure what we're shipping is needed. But all things point to an aggregate, we don't see a double ordering issue. Final point I'd make is we haven't been afraid to call those out. I mean, if you look at the Q3 of 2018 calendar, this was right when the trade war the tariffs were being announced.
And I don't know if you probably remember this, we had a 29% year over year growth in that quarter in our Marvell networking business and we were actually the 1st company to kind of point out, we think that this is not normal, right? And that we shouldn't expect this to continue and we tried to plan our business around that. So we've been forthcoming when we see it. At this point, we see the supply demand balance, at least in our view, to be pretty much in line, but we are paying close attention to it. Great.
Thank you.
Great.
Thanks.
Thank you. Our next question comes from Atif Malik of Citi. Your line is open.
Thank you for taking
my question and good job on results and guide. Gene, I have a question. You have talked about Avera plus Aquantia minus wireless adding about $100,000,000 to fiscal 2021 sales in the past. I don't expect you to break out these acquisitions anymore, but can you just talk about you're running above this run rate currently with Aquantia and Avera? Yes.
So we're very pleased with both acquisitions. I think last earnings call, we said that both Aquantia and Avera, they have been on track with our expectations. Definitely through this year, we think they are going to perform better than our expectations just based on the backlog, the order pattern, everything. So overall, very, very pleased with the 2 acquisitions.
Yes. And I'll just add to back that up. We had a very, very we had a very strong Q4 and Q1. And so the 6 months that we've owned both of them, we've gained a lot of confidence that not only were those the right acquisitions to do, but the level of top line contribution was going to be not only right where we thought, but probably given all the dynamics I mentioned with strength in cloud and strength in 5 gs that is benefiting certainly Alvera. And then on Aquantia, as I mentioned, we've had very good enterprise strength in some of these segments.
So we do see both of those probably performing better than we had indicated back at the time of announcement for calendar 2020.
Thanks. Thanks.
Thank you. Our next question comes from Harlan Sur of JPMorgan. Please go ahead.
Good afternoon and great job on the quarterly execution. On your DIY SSD controller with your lead customer, good to see the initial ramp this quarter. Given that this is a semi custom chip utilizing a vast majority of your IP, I assume that you're the sole supplier of this SSD controller. So in other words, your customer may be using different NAND memory suppliers, but they will be using your controller solution irrespective. Is my assumption correct?
Because I'm just trying to size the opportunity because your customers' platforms here has a historical cadence of ramping anywhere from 5 7000000 units in the 1st year to around 16000000 to 20000000 units by year 3. So pretty substantial opportunity if you guys are the sole supplier.
Yes. It sounds like you're right.
Yes. No, I'll take this, Gene, and then you can add. So yes, no, you're right. And in fact, the way to think about it is any all of these DIY opportunities are by design proprietary and sole source in nature. And they to your point, they typically incorporate pretty substantial and rich IP from Marvell.
And then as a result of that, we're able to command better ASPs and also add an additional functionality above and beyond what you would sort of normally think of a standard SSD controller as being valued at. So I think it's a great combination of any of these DIY efforts that we have going on and there are several of them where we have very close collaboration with the OEM to find really something that's very special and unique for what they need. And then the result of that is they can get pretty significant performance benefits across a wide variety of applications when it comes to flash based storage. So yes, proprietary, sole source and on track to ramp.
Great job on the execution.
Yes. I would just add, right, Matt said earlier, we already started production in Q2, which is much earlier than we happen.
Great. Thanks for the insights.
Yes. Thanks Arun.
Thank you. Our next question comes from Ross Seymore of Deutsche Bank. Your line is open.
Hi, guys. Thanks for letting me ask a question.
Matt, I want to go back
to the cloud theme that you talked about in your script. And kind of a clarification on a question, that greater than 10% number that you talked about, which is up a lot year over year, Just roughly, how is that split between the 2 segments, storage and networking? And more importantly, going forward, how do you expect that to grow? Customers can be very, very lumpy, especially on what I would assume would be your networking side of the business. It could be a little bit more steady on the storage side, but give us an idea of how we should think about monitoring the growth and kind of the lumpiness of that going forward.
Sure. Yes, thanks for the question, Ross, and I'm glad you brought it up. I mean, we definitely wanted to spend time on it. I think if you look at the last year or so, we've been very focused and certainly from an investor point of view, the demand for questions and information and the opportunity for Marvell, at least, has really been framed around 5 gs. And as I mentioned, that continues to be not only a great opportunity, we're actually starting to ship.
And so it's nice to see the results now catching up with the thesis. So very pleased about that. And so the thing that's really come together though in parallel as we did the various portfolio moves last year and even going back farther kind of the Marvell organic investments that we've made, We have been very focused on cloud and hyperscale. We do see that as a big SAM and a big opportunity. And so what's we thought it was an appropriate time to talk about that and break it out because a lot of things have come together and that business is now ramping.
And it's actually strong in both our networking and storage side and we see both of those opportunities as being quite significant. That's excluding ARM Server. I mean ARM Server, you just dump that one on top of it. But even ex ARM Server, we see the SAM for our networking and storage products and now the combined company with all the moves we've made as being an equivalently large opportunity as 5 gs. And while cloud spend can be lumpy, we certainly have that in other markets.
The carrier market isn't exactly the hallmark of stability. That thing certainly can move around as everybody that's been involved in base stations knows. So there can be some lumpiness to the business. But at the same time, because of the breadth of the products we sell into that market in cloud and even if you look at 5 gs, it's the same thing. Because we sell to a wide range of OEMs with the wide suite of technology, we do we are able to smooth out some of the lumpiness.
But I think the key thing I would point to is both businesses now are becoming significant for Marvell. Both businesses we expect to grow this year and beyond. And I think it's a horse race between the two on which one can be bigger if you look out over several years. And so that's really, Ross, why we wanted to spend the time on it. We think it's a pretty major opportunity.
But when it was so small a year ago before we did Aquantia I'm sorry, before we did we had Avera plus some of our own designs hadn't kicked in. We didn't want to do that. But it reminds me of when I was at Maxim and we had a very nascent emerging automotive business and then there was a point where we had an Analyst Day and Day and we decided to break it out because it had crested over 10% and then it was off to the races. So I think those are some additional considerations for you to think about on our cloud business beyond my prepared remarks.
Thanks, Ross. Can you have the next question, please?
Yes, sir. Our next question comes from the line of Blayne Curtis of Barclays. Please go ahead.
Hey, guys. Thanks for taking my question. I was
just curious, you provided a wider range. I'm kind
of just curious how you're thinking about that. Clearly, the factory shutdown that impacts you in storage,
it seems like most factories are kind
of back online. So it seems like some of the headwinds are tailwinds. I'm kind of curious what you're thinking about on the other side of the coin there as you look out into Q2 as well as the end of the year.
Yes. Sure, I'll give you both sides of the coin. I think on the one hand, you could take the argument you just said, which is what we see today, by the way, that things appear to be stabilizing and factories are coming online and countries are reopening and so forth. At the same time though, I mean, let's just take a big step back, right? The whole world and environment is highly uncertain and the rate at which this thing hit us, I think, got everybody off guard.
And so I wouldn't read too much into it other than we widened the range last quarter. And just out of the abundance of caution, there's no reason to tighten it at this point. But we don't see anything differently than you mentioned at the beginning. I mean, it seems like we've got our demand we've got our demand drivers. We've seen strong orders.
We've got good backlog position, those types of things. So we have confidence in our business, but it's really just around the broader level uncertainty. And I think our range, by the way, is very consistent actually with a lot of other semi fears. So I we'll probably leave it a little bit wider until we the world seems to stabilize.
Thank you. Our next question comes from John Pitzer of Credit Suisse. Please go ahead.
Yes, guys. Congratulations on the solid results. Matt, maybe another way to ask Blaine's question, you guys clearly have a different go to market model than a lot of other semis, more direct, less distribution. And at least relative to some of the broad based semis, you have a lot fewer customers. But I'm just kind of curious as I look towards the guide for July, it seems like a lot of your peers significantly discounted their bookings and backlog trends when they gave guidance for June, July.
Just trying to take a level of conservatism. I'm just wondering if you can give us any color on how you constructed this guide relative just to all the backdrop of the macro uncertainty? And as an example, some companies have guided to 100% backlog coverage where they normally guide to 75% to 80% backlog coverage.
Yes. Sure, John. Yes, happy to answer that. And I have seen that commentary. And I think it does have a lot to do with what you mentioned, in particular, probably the analog mixed signal guys who have a much broader customer set, highly distribution oriented consumer and industrial type of exposure and automotive as well where those types of markets seem to have been hit very, very hard.
I think our market mix is clearly benefiting us at this point. And our strategy that we really have been driving for the last several years, I think has become extremely pertinent in this new world. And so I think we're quite different than maybe some of the other broader based folks who I don't blame them for being extremely conservative and I understand those dynamics that they may face. But I'd say with us heading in, I mean, we also through the mix changes, because we're not selling as much in the consumer anymore as an example, the lead times that we get are solid. Our terms and conditions are quite solid.
The backlog coverage we get is especially with Avera coming in and think of it as Avera coming in and Aquantia and Wi Fi going out and you swapped out sort of consumer for infrastructure, you're just getting a much better backlog position. And on top of that, we've got the channels quite lean and we're also carrying some delinquency higher than we normally do. And so that gives which means we've we've got backlog that's out in Q3 that customers would like us to ship into Q2 and that has increased relative to where we normally run it. And so I think that those are some of the other data points John that give us comfort in the guide as well as just our starting backlog position has been higher than most quarters since we've been here.
Perfect. Great color. Thanks.
Our next question comes from Christopher Rolland of Susquehanna.
Matt, in your prepared comments, you spoke about Jensen and the DPU. And you guys were really pioneers in smart tech, NIC technology with liquid IO and Cavium there. And, you called that out specifically as a driver this quarter. I know Cavium used to have one key partner there that eventually went internal with their own development. But perhaps you can talk about this renewed interest.
How broad is it? What kind of levels are we at? Are we close to old levels? And is it going to be sustainable in your opinion this time?
Yes, great. Thanks, Chris. Yes, great question. So first point on BPU, I thought that was really, really well done by our colleagues over at NVIDIA and it certainly resonated. I think that was the clearest way that I've seen anybody explain what these products do.
And I think the market in general has gotten confused over what's a smart NIC, isn't a NIC kind of a commodity thing And there's terms like fungible and liquid and it just I think this is a much crisper way to describe it. And so and it certainly fits, as you point out, kind of a key IP and a key area that Cavium really pioneered. And so with respect to our SmartNIC offerings, I think kind of consistent with the question that Ross asked, we spent a lot of time talking about 5 gs last year, but very quietly we've been continuing those roadmaps. And so that is ramping. It's highly sustainable.
We and we're engaged with multiple OEMs between the ramps that we're getting plus new opportunities and engagements. And I actually think we're in probably one of the best positions of any semiconductor company to truly tailor and optimize solutions based around a smart NIC type of architecture, because we have all the pieces, right? We have the latest generation OCTE Encore, which is highly programmable. We have network connectivity in Ethernet. We have security.
We have access to and development and investment in the most advanced node process technologies. And we have ASIC capability, which is a trend that we see as well where some of the large hyperscalers are just saying, let's go skip over a liquid IO type of solution. Let's go directly to our very special thing that we want. And we think we can provide a similar type of solution there just like we do in our partner model on baseband, where we contribute a big chunk of the IP. The customer can also contribute.
We stitch it together. They get exactly what they want. We can do it very quickly without the customer having to hire a gigantic design team and take 3 years. So I think it's a very compelling opportunity, Chris, both on the current ramps we're seeing on the latest generation products as well as our roadmap and the ability to customize and drive this product area forward. Thanks for the question.
Thanks, Matt. Congrats on the quarter.
Yes. Thanks. See you.
Thank you. Our next question comes from Joe Moore of Morgan Stanley. Your question please.
Great. Thank you. You talked about Thunder and the enthusiasm that you have there. Can you talk a little bit about some of like what are the key workloads that you're addressing? And is this I think at one point there was a thought that this sort of is a replacement for Intel on a broad base of applications and versus something that's kind of accelerant for a few key workloads.
Like what's just the framework of how you're thinking of that now over the next 3 to 5 years?
Sure, Joe. Yes. And I think the world the whole funder investment that we're making has continued to be really focused down. I mean, if you went back to sort of pre acquisition of Cavium, I think there was a view that said there's a very large market for this. We can sell to a whole ton of customers across a whole bunch of different applications and serve a number of workloads and go head to head with more general purpose typex86.
And really the direction that we've taken is to really focus the product down and tailor it really around a handful of key vendor key customers in the cloud area. And within that, as an example in our announcement publicly we had with Microsoft, we're the first property that we're in is in Azure storage. And the way we developed and defined the part was really ideal for that application. But we can certainly do a high level of customization for these hyperscalers because they know exactly what they need. And so we do see opportunities in expanding into more compute oriented applications and workloads, but that's sort of in the future.
But again, I'd sort of transition the Thunder story just to bridge everybody from, hey, it's broad based, it's going to go take on X86 and take on the world across all kinds of applications to really the evolution over the last few years under a more focused effort to really focus around the hyperscale opportunity and tailor it specifically to those customers and what they need.
Great. Thank you.
Yes. Thanks.
Thank you. Our next question comes from the line of Tore Svanberg of Stifel. Your question please.
Yes, thank you. Congratulations on the results and the new branding. I really like the logo. Matt, not to sort of steal your thunder from the Analyst Day in October, but you talked about in the automotive space, you kind of see the zonal architecture that's very similar in the sort of enterprise networking world. I was just wondering how that's going to play out.
Will you be working with networking companies here or will you be working with the hyper guys or the auto companies? So just trying to get a head start on how that whole world is going to look like. Thanks.
Yes, great question. Here's the way I'd say it. It's very much with the automotive customers themselves, and it's very much with the OEMs themselves. And a little bit like if you go back in time, the automotive OEMs probably in the mid-2000s to kind of 2010 era started taking a lot more control over their semiconductor strategies. And that was really generally around things like infotainment, EV, ADAS and you also started seeing these gigantic automotive OEMs with full semiconductor teams.
First, it was sort of purchasing and quality, then they started adding design capability and resources and they've built up over the years quite a capability. So they take very seriously they want to control their own destiny when it comes to their future car architecture. So we have a huge head start because of our Ethernet opportunity where we're designed in. We've talked about this a couple of times. We're in 16 different OEMs are going to use are using or are going to use our Ethernet solutions to provide in car networking.
And that provided a nice beachhead to go in and also show them all these other types of IPs we have and now we're getting traction there. And so this is a longer term opportunity, but it can be quite significant in nature because I would the analogy I would give you would be, I think the sea tide shift we saw in automotive in the kind of mid-2000s in sort of the infotainment revolution that benefited all kinds of companies, all kinds of semiconductor companies. The new wave of automotive suppliers came out of that. One was my prior company, but also many others had huge benefit on the infotainment revolution. I think this data centric evolution in automotive is going to be just as substantial and really plays to our strengths.
And so we look forward to articulating this in more details at our Investor Day. This is a long haul type of opportunity, but we've got enough traction now that it's going to be worth calling out and explaining to investors and the analyst community what we're up to and what the opportunity could look like, especially for our long term investors who want to who are now asking me, hey, this is all great with what you're doing in 5 gs and what you're doing in cloud, what's next? And so we think there's a nice story to be told there about what's beyond these other cycles that can drive the company for a long time in the future.
Thank you for that.
Thank you. Our next question comes from Sareen Pajjuri of SMBC. Your line is open.
Thank you. Thanks for squeezing me in. Matt, I just want to go back to the 5 gs topic. You alluded to Q3 being potentially Is it Is it what's giving you confidence? And it looks like it's mostly coming from China, your Q3 commentary.
Could you also talk about what's going on in some of the other regions like U. S. Or Japan and Korea? And also touch upon India because that's been a big market for Samsung historically. Thank you.
Yes. Thanks for the question. So a couple of things. I think one is, given the ramp that's coming and the lead times associated with these products, we have strong backlog today and we have strong bookings in this area, both in Q4 as well as in Q1 that certainly give us confidence that Q2 and Q3 are going to look very, very, very good in this area. And the reason at least that we're hearing and that we can infer both from looking at the outside market outside in as well as from our customers is that primarily I'd say primarily driven by China.
But I think the OEMs are all gearing up. To your point for the other markets. We see that Japan will start to roll out later this year. There was a concern that that would get pushed or paused completely due to the Olympics moving out and then COVID-nineteen. But it looks like that you well may not be as robust and we don't know yet on any of these things whether it will what the slope will look like.
But Japan is certainly talking about rollouts starting later this year. The U. S. Carriers, once they can start putting people back out installing things, they're planning on a second half in more robust deployments. And then India, they've got and the U.
S. By the way, both have pretty important spectrum options that need to get completed, we think, to really drive 5 gs. As you know, in India, we still have a strong 4 gs presence to our lead customer, but really it's 5 gs that's going to drive the growth. So we these things may be moving around a little bit and they may be moving out, but it's definitely a when, not a if. And I think what's nice is it's hard to handicap these, but what I'd say is even though some of the other one, if you looked at where we started the year, some regions are probably pushing out a little bit more than what we thought, but then the China opportunity actually sort of dwarfs that and offsets it.
So I think net net, we do see countries that are going to roll this out. It's going to take time, but it's encouraging to see at least a large deal really driving this forward. And then ultimately, that's rippled back to orders on our books that we need to go fulfill. So appreciate the question.
Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.