Good afternoon, ladies and gentlemen, and welcome to our conference call announcing Marvell's acquisition of GlobalFoundries' Avera Semiconductor business. Today's program is being recorded. After the speakers' prepared remarks, there will be a question and answer session. I would now like to turn the call over to Mr. Ashish Saran, Vice President of Investor Relations for Marvell. Please go ahead, sir.
Good afternoon, everyone, and thank you for joining us, particularly on such short notice. After the market closed today, Marvell announced its planned acquisition of GlobalFoundries' Avera Semiconductor business. To discuss this announcement, I'm joined on the call by Matt Murphy, Marvell's President and CEO, Raghib Hussain, Marvell's Chief Strategy Officer and EVP of our Networking and Processes Group, and Jean Hu, our CFO. A press release on this transaction is available on the investor relations section of our website at www.marvell.com. This conference call is being webcast live and a recording will be available via telephone playback and also archived in the investor relations section of our website. As a reminder, today's call will include forward-looking statements regarding our future business performance, the intended benefits of the acquisition, and the expected timing and completion of the proposed transaction, as well as the financial impact to Marvell.
These statements include risks and uncertainties that could cause our actual results to differ materially from the statements made on this call. Please refer to our press releases today and our recent filings with the SEC for information on specific risk factors. Finally, comments made during today's call will primarily refer to non-GAAP financial measures. To quickly summarize the transaction. Under the terms of the definitive agreements, Marvell will pay GlobalFoundries $650 million in cash at closing, plus an additional $90 million in cash if certain business conditions are satisfied within the next 15 months. We intend to finance the acquisition by accessing the debt markets and plan to close by the end of fiscal 2020, subject to regulatory approval as well as customary closing conditions. We expect approximately $300 million in revenue from the Avera business in the first full year post-close.
If certain business conditions are satisfied, triggering the $90 million incremental deal consideration I just mentioned, we would expect at least an additional $40 million in revenue from the Avera business over the same timeframe. In either case, we plan for this transaction to be accretive to our non-GAAP earnings per share in the first full year after close. I will now turn the call over to Matt for his comments.
Thanks, Ashish, and good afternoon, everyone, and thank you for joining us today. Our acquisition of GlobalFoundries' Avera Semi business represents a significant step in Marvell's strategy to become a global semiconductor solutions leader for the infrastructure market. With Avera, we are creating an infrastructure ASIC powerhouse with complete product design flexibility, leveraging advanced technology platforms and global scale for strategic wired and wireless OEMs. Stepping back, as we outlined at our investor day last year, Marvell's current product portfolio encompasses standard products and semi-custom solutions, which we refer to as our partner model, where Marvell provides the majority of IP and customers add their own unique capabilities or features. We also identified a third category, full custom ASICs, which is Avera's core business.
For decades, they have been a leading provider of highly complex custom designs for the world's leading infrastructure OEMs, and their ASICs are at the heart of critical enterprise networking, carrier, and data center applications today. Previously, as part of IBM Microelectronics, Avera's world-class engineering organization successfully developed more than 2,000 complex designs over its 25-year history and built a significant business supported by approximately 800 very experienced technologists. Avera brings highly innovative design competencies in analog, mixed-signal, and SoC development and a rich IP portfolio underpinned by approximately 800 patents and applications. They have long been a leading provider of high-speed SerDes for network infrastructure, have developed performance and density-optimized embedded memory, as well as advanced packaging capabilities across their product portfolio. Avera has built long-term partnerships with market-leading OEMs in wired and wireless infrastructure.
They provide the critical technology inside switches, routers, wireless base stations, and other key infrastructure applications. More recently, they have begun to address new applications in next-generation cloud data centers with multiple products in development today. Marvell is a leading supplier of standard and semi-custom products into these same markets, and we have continued to expand our own pipeline for custom solutions, leveraging Marvell's leading IP and technology platform. For example, in 5G base stations, we have seen an increase in customer engagements to develop new custom products to complement our wide range of basebands, processors, switches, and PHYs. These new opportunities are driven by our customers' growing needs for power and cost-efficient solutions to address their broad portfolio. Many of these are also designed to replace less efficient and more costly FPGA implementations.
Avera has been benefiting from this same dynamic, delivering optimized custom digital front-end solutions for wireless radio heads in 4G and have design wins in 5G as well. These solutions further expand Marvell's addressable market and illustrate the broader opportunity for custom solutions in 5G applications. Altogether, it's become clear that our opportunity in the 5G market is substantially larger than what we've previously outlined. These 5G examples are indicative of the broader opportunity across the infrastructure market. Our customers need to cost effectively and quickly build highly differentiated solutions in an era of rapidly rising design costs and faster time to market windows. They increasingly want a partner who can enable varying degrees of customization, mixing and matching the best combination of IP all the way to a full custom ASIC to best address their unique architectural needs tailored to their system-level requirements.
Through our Avera acquisition, Marvell will now offer customers an unparalleled breadth of standard, semi-custom, and full ASIC capabilities all under one umbrella. As we fully integrate and align the Avera business, we expect to pursue additional strategic ASIC and semi-custom engagements with tier 1 customers in our core markets for networking, data center, cloud, and automotive applications. We believe that this represents a multibillion-dollar increase to Marvell's annual TAM opportunity. The addition of Avera's talented team and extensive custom design expertise will accelerate our ability to capitalize on these opportunities and capture substantially more content across the broader landscape. We are also looking forward to furthering our successful partnership with GlobalFoundries.
In closing, our acquisition of Avera adds world-class custom chip design capabilities at a key market inflection point, will enable us to pursue additional 5G opportunities, and at the highest level, represents a major step in Marvell's transformation into a global leader for infrastructure semiconductor solutions. Operator, let's open the line for questions.
Certainly. Ladies and gentlemen, if you'd like to ask a question, please hit star, then the number one key on your touch-tone telephone. If your question has been answered or you wish to remove yourself from the queue, you may hit the pound key. In the interest of time, we ask that you please keep your questioning to one question and one follow-up question before rejoining the queue. Again, that is star, then one if you'd like to ask a question. Our first question comes from Vivek Arya with Bank of America. Your line is now open.
Thanks for taking my question. Matt, you mentioned new opportunities on the 5G radio side and at cloud data centers. Can you give us a sense for what have been Avera's kind of historical revenue and margins? Typically, in these ASIC opportunities, the revenue side is fine. The margin side is a little bit below what your business model has been. What has been kind of their historical revenues and margins and what kind of growth rates and synergies are possible 12-24 months after closing the deal?
Sure. Yeah. Hi, Vivek. Great question. Yeah, as we just mentioned, the businesses are approximately $300 million in revenue. We anticipate next year after we close the gross margins of the whole portfolio are around 50%. The way we think about it is the pipeline of opportunities that we have as Marvell as well as Avera, that we're going to deploy the resources towards, are more in line with the company average. Think about this as the existing book of business comes over at a certain level of margin, and then obviously as we add value and we address some of these higher-end applications, we think that the new opportunities will be accretive to that gross margin. That's the way to think of the trajectory of this business.
Got it. In terms of just the exposure, can you give us some sense of the specific exposure to China or perhaps of Huawei, just because it's in the news so much? If you could also kind of give us what Marvell's organic exposure is to China and Huawei, just to give us an idea.
Sure. Yeah. Happy to do that. Yeah, this is certainly an interesting time to be announcing this. Let me give you some context. For Avera, the primary agreement that we've called out, the $650 million agreement, the exposure is quite minimal. I'll get this question later, so I'll cover it now from a regulatory point of view, that agreement does not require China approval, and so therefore the exposure to China revenue is quite small. On the Marvell side, the way to think about this is the Huawei business, you should think about as being sort of mid-single digits percentage of our revenue. All of you know, we have an earnings call next week, and so what we'll do is we'll comprehend, as much as we know at that time, the impact of the ban into our guidance at that time.
Just give us a week in change, and we'll do the guidance then. To just give you the number up front, it's about mid-single digits.
Thank you.
Thank you. Our next question comes from Harsh Kumar with Piper Sandler. Your line is now open.
Yeah. Hey, guys. First of all, congratulations. Matt, I went on the avera website, and it basically says revenue's in the $500 million range. I'm wondering if that's old information, or are there parts of the business that you are probably or likely not bringing over? Could you maybe explain that? My second question, I'll ask it and then get back in line. You recently said you're looking to acquire Aquantia, I think, and now Avera. Is there a change in the philosophy that is happening, or are these just opportunistic kind of things that are falling your way and you feel like they're good add-ons and they don't require China approval, so minimal headache? Could you just maybe clarify that for us?
Sure. Great two questions. On the first one, yeah, the way to think about it is not all of this business is coming over. That's the first point. There's some of the business that will remain. Think of that as sort of being maybe end of life or just programs that don't come over. The second is, as I mentioned in the primary agreement, there's revenue that also doesn't come over. Obviously, what we said was, if there's this option agreement, then there's additional revenue that comes with it. Yeah, we've looked at the ins and outs, and that's the bridge that not all the revenue comes over. What we're left with, we think is pretty clean. Of course, with the team and the pipeline, we can grow it from there.
On your second question, yeah, we certainly are doing two of these relatively close to each other. A couple of thoughts on that. One is, in an ideal world, right, you could time these the way you wanted to time them. They are close together, but the way we think about it is, as you said, first, from a regulatory standpoint, neither requires China, which simplifies things. The second is that on a relative basis, Aquantia's a relatively small deal in terms of the revenue and just the complexity and the number of part numbers that come over. The other way to think of it is they're actually two different businesses within our company. When you think about sort of the teams that need to spin up and integrate, the R&D folks are actually different business units in the company.
I feel very good about the capability that we've now put together in Marvell. If you look at just the Cavium integration, just an update on that, we did the ERP cut-over two months ago. That's done. That was done actually ahead of our own schedule. We built a very strong operational capability. We feel pretty good about doing both of these, and they both make sense, and they happen to just be a fairly short period apart from each other.
Congrats, guys. Thank you.
Yeah. Thanks, Harsh.
Thank you. Our next question comes from Gary Mobley with Wells Fargo. Your line is now open.
Hey, guys. Good afternoon. Thanks for taking my question. With the pending acquisition of Avera and Aquantia, can you go over the cash flow assumptions for the balance of the year and how that impacts what the debt to EBITDA ratio will be coming out of the close of both these acquisitions?
Sure. Let me have Jean answer that question. Yeah, go ahead, Jean.
Hi, Gary. Thank you for the question. As you know, our business continues to generate a strong cash flow between signing and closing of both Aquantia and Avera transactions. We think at the closing time, we probably need to borrow, if we need to borrow the whole consideration of this transaction, which if we consider the $740 million whole consideration, our gross debt to EBITDA ratio at closing before any integration, everything, it's 2.5 times. Our net debt to EBITDA ratio is only going to be two times. After closing, we'll certainly focus on reducing the debt level with a strong cash flow we'll generate from the combined three businesses. We expect to take down the debt quickly to below the two times of gross debt to EBITDA ratio.
Okay. Helpful. Just as a quick follow-up. Since this is a custom ASIC business, I presume there's a fair amount of customer concentration. Can you cover that topic and speak to the degree to the customer concentration?
Yeah, sure. Without going into all the details, what I'd say is, they've historically had a number of significant customers. That being said, they also have a number of new customers that they've engaged as well. I would say the way to think about it, rather than the customers is they have a pretty good blend of sort of wireless infrastructure business, AKA base stations, networking business, think switches and routers and enterprise, and then some emerging applications in the data center. I'd say of those, the primary one has probably been in the wireless market.
Okay. Thank you, everyone.
Yeah. Thanks, Gary.
Thank you. Our next question comes from Ross Seymore with Deutsche Bank. Your line is now open.
Hi, this is Jean for Ross Seymore. Thanks for letting me ask a question. Jean or Matt, can you talk about any cost synergies? You discussed that not all of the business from Avera is coming over to Marvell. Will all 800 employees come over as well, or what is the headcount coming to Marvell?
Yeah. I'll answer, Jean, you can add, too. The way we think about this is just what's the spending going to be on this business when it comes over, right? The spending is going to be about $100 million. It's not a typical deal where you say, well, here's what they were before, let's just start doing the math. Just think of it as it's a $300 million revenue, 50% margins, $100 million of OpEx. Obviously we get operating income accretion. We're going to work out the entire integration plan on how we combine the Avera team with the Marvell team and put together what we think is going to be a best-in-class ASIC organization.
Okay. Thank you. GlobalFoundries has discussed not supporting the leading-edge nodes. What will Marvell's manufacturing strategy be for this business going forward?
Sure. We're actually very excited about this, I think certainly we're excited, the Avera team's excited, I think the customer base is going to be thrilled. The reason I say that is that the plan is for, obviously there's commitments and there's projects in flight that Avera's working on. We're very excited about those. Those are all primarily in GlobalFoundries. For the new engagements where we need to really access the leading edge, by the way, Marvell, as you can imagine, is already heavily investing in this area anyway in terms of setting up our 7-nanometer platform and then eventually 5-nanometer platform.
The way to think about it is you're going to take this very talented Avera design team, couple it with Marvell's technology platform at 7 and beyond, it really gives the Avera design team and our customers access to the leading edge, which is our primary partner has been TSMC. If you think about TSMC process technology plus really Marvell's very rich IP portfolio, plus the design team of Avera, we think it's a very compelling combination for customers that want to participate at the advanced nodes. That's our strategy.
Great. Thank you.
Thank you. Our next question comes from Harlan Sur with J.P. Morgan. Your line is now open.
Good afternoon. Congratulations on the acquisition. I think the Avera team has a great track record of advanced custom ASIC designs. Matt, in response to the prior question, if you look at most of the advanced ASICs currently being designed, they're on 7 nanometer and a few on 5. I guess the question is, when you answered the prior question, does Avera already have 7 nanometer and 5 nanometer qualified ASIC design flows at any other foundries? If so, what specific foundries?
Sure, Harlan. Great question. Here's the way to think about it. Last, call it summer, right, or in the early fall, GlobalFoundries made the strategic decision at their company level to stop investing in the bleeding edge. Made a strategy pivot and they've announced all kinds of very, I think, positive moves since then to really reinforce their strategy. Part of that was by not having the leading edge, obviously, their customers needed a solution. That's where Avera Semi was born. This team has already for, since probably more than six months, since last fall, let's call it, engaged with TSMC. They've actually won designs already there. They have active programs with TSMC today. They're well underway, as are we as well. I mean, we've got our first tape out coming in 7 nanometer.
I'm talking about now, this is all 7. Our first tape out in Marvell is coming July timeframe. The two of us are, I think, well-aligned, and they already have a running start. That's the first point. The second is, you're right that the world is also moving to 5. We've also begun work in that area, and the way I think about it at a high level strategically is if we're able to leverage more designs right on each of these process technology platforms with the combination of the Marvell business plus the Avera business, it has all kinds of benefits, right? It has scale benefits. We can leverage the IP across many more tapeouts. It helps our cost structure. There's all kinds of benefits. That's another reason to do this was quite frankly, we've got a big lift, right?
To go do all the things we need to do at 7 and 5. Now we've got access to a whole multi-billion-dollar market, where I think there's going to be significant demand for our products. That's the way to think about how Avera's team has made the pivot, and I give them a lot of credit. They had designed in a GlobalFoundries IBM environment, and they've really done a good job of accelerating the transition to advanced node, and we look forward to working with them and hopefully with the Marvell team in place, actually accelerating their progress.
Thanks for the insights there. On the press release, you discussed a number of radiohead ASICs that will be deployed into a leading wireless networking OEM. Are these mixed signal ASIC chips, in other words, digital, analog, and RF blocks all on the same chip? When do these first ASICs start shipping? Thank you.
The products that they're working on, they have all these mixed signals saying it has an analog aspect and the detail which is typically in the digital front-end solutions and RFICs in this. Avera Global Semi actually has been engaged in those kind of design for several generations, and some of the design that is in the current as well as some of the early design ones, as Matt mentioned in seven nanometer, is in that category.
Great. Thank you.
Carl.
Thank you. Our next question comes from Christopher Rolland with Susquehanna International Group. Your line is now open.
Hey, guys. Congrats. If I heard the presentation correctly, maybe I'm reading too much into this, but is the idea that you guys have ASICs for the baseband unit and baseband processing, and this has more of a remote radio head focus, or is this also for the baseband processing?
Sure. Let me up level it, and then I'll go 1 level down for a second. Again, just to take a big step back, right? There's two big benefits in this combination, right? The first is, as we've just discussed, all kinds of benefits in Marvell fully entering the ASIC market, right? We discussed all the different reasons why that's good. The other piece of it is that it enables us pretty significant additional design capacity to go after, quite frankly, a growing backlog of opportunities we have across a number of OEMs for base station 5G silicon. What we've seen is the demand we have and the opportunities, they actually span the existing business we have in the control and transport processing, as well as the baseband.
As Raghib mentioned, with Avera, they have expertise and a position in the digital front end, which is up in the radio head. That's why we think this expands our SAM, and we actually see opportunities now as Marvell and now with Avera, in the radio head and in the baseband unit across multiple digital technologies and mixed signal technologies that are critical for the next generation of base stations. Yeah, I think from a positioning point of view, certainly it expands us into a new segment that we haven't historically been in.
Great. As I think about customers. Okay. Yeah, thanks, Matt. As I think about customers, the engagement here is perhaps a little bit different than your leading customers, if I understand that correctly, and maybe talk about that. Also, the data center opportunity, what are these products? Are they 100 gig switches? I'm just unfamiliar. Thank you.
Sure. Yeah. I think on the base station one, what I'll say is I think we are pleased that this gives us some higher exposure to some additional customers that we didn't have as much exposure to. I think that's one of the benefits we see with this. Obviously, we've got a lot to offer to the market, to the extent we can be more broadly positioned, I think that's a good thing. On the data center side, I think we're not at liberty to talk about the details of what's the individual applications. Those are fairly sensitive to those particular customers.
There is multiple projects in flight, certainly as Marvell, we've gotten numerous requests over the last couple of years here and even recently to, "Hey, can you guys participate and help us in some of these things?" Which, as you know, there's a whole bunch of very customized data center cloud silicon that needs to get built. We think that we can go back in now having this capability and probably be more aggressive there.
Thank you, Matt. Congrats, guys.
Yeah. Thanks, Chris.
Thank you. Our next question comes from Ambrish Srivastava with BMO. Your line is now open.
Hi. Thank you very much. Matt, I'm not sure I heard you answer this question Vivek had asked earlier. If I missed it, I apologize. What has been the revenue trajectory for this business over the last three to five years? Then I had a follow-up, please.
Well, what I'd say is that right now the business is doing quite well. The last couple of years it's been growing. There was a period of decline if you go back years and years ago, that transition from IBM to Global, and they had to kind of reinvent themselves and find additional customers and leverage sort of the GlobalFoundries footprint. Today that business is certainly from 2017 to 2018 and 2018 to 2019 is on a growth trajectory over the last couple of years. Certainly we're hopeful that when we close and in calendar 2020, we'll have a baseline of business that will grow on its own. Then obviously we would add these incremental opportunities on top of it.
Okay, that's helpful. Then on the accretion and the gross margin front, this is what I was struggling with. ASIC businesses inherently are lower margin businesses. The accretion is going to come immediately off the bat. Is it going to come from the OpEx side? Then why should gross margins go up to Marvell margins? It's not like these guys were taking lower margin businesses. What's the path to get the 50%-60%? Not sure I quite understood that part. Thank you.
If you look at the current, as Matt said earlier, the current business itself is in 50 range. As we are utilizing the team going forward, we are actually using the team in several projects, not pure ASIC only, but also on the CSSP side, where we have demands right now. As we add more IPs, our own IPs in the overall solution, this is how the overall margin of the business increases.
Yeah, that was a great clarification. I think, again, we're not going to get beyond this conversation into super micro detail on exactly where these are. Think of it as the team that is coming over will continue to participate in pure ASIC. That has a certain margin profile. As you mentioned, it's never going to be as high as if we were adding value, but the team itself is going to be working on both. The economic value derived from the team's engineering, the investment we make will blend up to a higher margin. I think that's the way you should think about it. It'll be a combination of both.
Okay. That makes sense. Thank you.
Thank you. Our next question comes from Quinn Bolton with Needham. Your line is now open.
Hey, Matt. You talked about a substantial library coming over. Will you guys have to port that over to TSMC, or would you just look to develop new IP blocks for the five and seven nanometer node? A second follow-up question. You mentioned some of the high-speed SerDes capabilities. Next generation switches for the data center look to be moving to 112 gig per lane and wondering if you guys acquire that capability with this acquisition. Thanks.
Sure. I'll take the first part. I'll have Raghib do the second on SerDes. Think of it this way, we have our own seven nanometer readiness plan, right? Avera will be immediately able to benefit from that. There's no porting that needs to come over. For seven and for new designs, obviously we'll benefit from their learnings, right, as they come in. Think of it as we're already doing that today, so that's a leverage point. On SerDes, maybe Raghib, you want to comment about it.
Yeah. To clarify, there is no such porting is needed. Avera independently as they have started working for the last three quarters working with TSMC on the seven nanometer. Certain IPs, they are already developing on seven nanometer. Independent of that, Marvell itself is developing seven nanometer our own IPs and including the 56G and 112G SerDes. Our first seven nanometer tape out is in two months. We have already been investing in those areas. We have those IPs as well as Avera. Our plan is to, when we combining these IPs overall IP portfolio, that will be our offering for the standard product as well as for ASIC.
Yeah, I would just add to complete it, I think that was for seven. Just if you go back one node, even at 12, that's going to be our 56 and 112 is already underway, and it'll be ready much earlier. From an Avera point of view, you may know this, but that team has a very strong reputation in their SerDes capability. In 14 nanometer today, they have 56G SerDes that's in production as well. Rich set of IP and very talented SerDes teams, this is a scarce resource in the industry, and I think we're fortunate to be able to have two very talented teams come together. Just to be very clear, again, we have 112G, 12, we're going to have it in seven and beyond.
Hey, Matt, can I ask one quick follow-up? Does Avera also have a multi-gigabit analog to digital and digital to analog converter capability that might come into play in those remote radio head ASICs?
Yeah. I think what I'd leave it at without kind of drilling quite yet into all the individual pieces we get. I think between the Avera team and the Marvell team, I just leave it at we both have very strong mixed signal capability, and certainly there are very strong converter people in both companies. I think I would leave it at that at this point. We can share that later in terms of when we get deeper into our roadmap discussions and how all this IP comes together and ultimately our vision to really have a meaningful role in the entire signal chain once it comes off the antenna into the digital domain. I think there's an exciting future there. Let's share it at the right time.
Okay. Thank you.
Thank you. Our next question comes from John Pitzer with Credit Suisse. Your line is now open.
Yeah, good afternoon, guys. Matt, congratulations on the announcement. A lot of uncertainty around the Avera asset over the last year, just given what GlobalFoundries has been going through. I'm just kind of curious, as you were doing the due diligence around the revenue ramps, to what extent is this sort of still Avera needing to hit technical milestones? Or to what extent was this just uncertainty about long-term roadmap viability and scale that becoming part of Marvell kind of gets rid of? I.e., how quickly do you think some of these revenue opportunities may materialize?
Sure. I think it's two pieces. I think one is on the materialization, we certainly have a number of projects today that we could close it and get them started. We would do it as soon as we could. I think that part we feel pretty good about certainly our pipeline and backlog. I think on the revenue coming over, I think the team and even myself, I think we've done a very thorough job to understand the moving pieces and de-risk some of that. What I would say is in the revenue that we're planning on for calendar 2020, think of it as existing blue-chip proven customer base that we understand the dynamics of in terms of if it's ramping or it's declining and why it is. I think we've gone through that in a pretty detailed manner.
That's not contingent on a lot of new ramps or things happening or smaller customers ramping to be huge customers. Think of it as a continuation from a trajectory point of view of where that business has been going.
Then, Matt, as a follow-up, is the real opportunity here just displacing what's being done on FPGA today? Or I guess who are the other competitors here that we should be thinking about?
If you look at the overall real opportunities in terms of the output in 3 categories, the data center, the wireless infrastructure, as well as the high-end networking market. As you all aware that the data center folks have a lot of chips that they make their own. It's a big opportunity over there to expand in that business area. Similarly, there are opportunities to get into the high-end networking ASIC solution. The wireless infrastructure itself. Some of them, of course, in the wireless infrastructure side are in FPGAs, and traditionally they do convert into ASICs, some of them will overlap of that. There are others which has been
Which has been the ASIC itself, that will also be. On the wireless infrastructure side, it is more on the one which is overlap over, to convert from the FPGAs to the ASICs.
Thanks, guys.
Yeah. Thanks, John.
Thank you. Our next question comes from Craig Ellis with B. Riley Securities. Your line is now open.
Thanks for taking the question, and congratulations on the transaction. Matt, I wanted to start with you, but I wanted to start higher level. If we take a step back and look at this transaction, Avera Semiconductor, and if we look at Aquantia, can you just talk about how satisfied you are with the networking segment's technology and product roadmap completeness once you get those two deals into the portfolio? What other needs would you have, if any?
Thanks, Craig. To start at the high level, clearly, as we outlined with Aquantia, we're strategically very excited about the automotive opportunity, and I think certainly the multi-gig area was very complementary to what we were doing. That certainly from a, you ask how satisfied are we, I think that definitely fits the bill because we have a lot of conviction that that transition to Ethernet networking in automotive is going to happen. The fact that, Aquantia and us fit so well together there, I think is huge. That definitely filled a need, and I think it's going to be a home run for our customers as well.
The exciting thing about Avera is it really fills what I call a need that's been there for some time, but we've just not to really staff up and build an entire end-to-end ASIC operation organically, it would take forever and be very challenging. The fact that we had the opportunity at this time to go acquire it and get to work with the team, I think, just worked out really well. If you just think about the arc of where we've been, right? Coming up on my three-year mark here, right? From where we were sort of three years ago to today, we just continue to really march forward with our transformation and this company to make it a viable, healthy, extremely competitive supplier to the infrastructure market.
I think if you're going to be big in the infrastructure market, you really need to also be able to offer this customized solution, especially based on some of the trends that I think we're all seeing. I think we all read the news and see that various hyperscalers around the world are all investing in their own teams to develop custom silicon. All of those companies need a partner like an Avera to actually go make those designs manifest themselves into reality. Lay them out, yield them, ship them in volume. They need to partner. I think, again, the combination of having that expertise, which Avera brings, because they really know how to do that well, combined with, as I mentioned, our Marvell Technology platform and just our total focus on this market, I think it's going to resonate really well.
We feel like we're going to be in a very good position when we are able to close these transactions, bring the teams in, integrate them, and present a very holistic, compelling value prop to our customers.
Thanks for that, Matt. The follow-up's for Jean. Jean, as we think about the two deals together, can you address how confident you are getting the business back up to the target margin model? As we look at the amount of investment that's coming into networking, do we think any differently about the intensity with which the company will be investing its R&D in the storage side of the business? Thank you.
As we discussed, both deals are accreative to our non-GAAP operating income and also non-GAAP EPS after closing. Actually, by doing two deals, we continue to increase our scale, especially in the networking, the core area we're investing as a company. I think going forward with the integration of two deals, you're going to see, of course, our top line revenue to be continuing to expand. On the gross margin side, this transaction, as Matt mentioned early, is about 50% gross margin at the very beginning. Over time, it's going to migrate up with the mix change, with the new designs we're going to have. The most important thing is the operating leverage. We have the infrastructure to really integrate the two deals and to enhance our capability and also generate more free cash flow and more operating income.
It only helps us to achieve our target model going forward.
Thank you.
Thank you. Our last question comes from Ruben Roy with MKM Partners. Your line is now open.
Thanks. Matt, if I could follow up on Craig's question, the last part of it. I was wondering, there's a lot going on with infrastructure right now, and just wondering how you're thinking about some of the core Marvell businesses. Has anything changed with the way you're thinking about investing in some of the core areas like storage or connectivity? Thanks.
Got it. No, it's a good question. I mean, again, we continue to be very bullish on our networking segment. As you've all seen, storage is going through what it's going through. Our thesis in storage very much remains the same in terms of our focus. At the same time, look, we're disciplined operators here, and we look at our investments constantly. From a strategic point of view, we're still very comfortable with our storage strategy. We like it, but at the same time, that business has been down for all the reasons we've discussed. I wouldn't signal any change there at this time.
Okay, thanks. Just a quick final follow-up. Has Marvell or Cavium worked previously with Avera in any aspect or been involved in reference designs or similar customer design wins or anything like that?
Not to my knowledge, no.
Okay. Thanks, and congrats.
Yeah. Thanks, Ruben.
Thank you. That does conclude our Q&A portion for today, and that does also conclude our conference. Thank you for your participation. You may all disconnect. Everyone have a great day.