Marvell Technology, Inc. (MRVL) Earnings Call Transcript & Summary

Hi, everybody, and welcome to the last session of our conference here, and it's going to be a good one. We're joined by Matt Murphy and Ashish of Marvell. Thank you very much, gentlemen, for joining us.

Yes. Thanks, Chris.

Yes. Thanks, Chris.

So we had some results last night. We had a very interesting quarter. I'll talk about that a little bit in a second. But Matt, I was wondering if maybe you could just give us a tiny quick summary particularly regarding maybe your outlook. I've had a couple of questions from investors how you're feeling about the year, how you're feeling about that 30%-plus number. And just anything else you want to add?

Yes. Great. Thanks, Chris, for having Ashish and I on. And it was a great quarter. If you look at all the numbers kind of across the board, right, executed well in all the markets. Earnings per share up pretty massively year-over-year. The next quarter guide also implied up like 70%, right? So not too bad, record margins, record revenues, record operating margin. It was an outstanding quarter with more to come. On the question on the outlook, yes, I did want to clarify that. I think that may not have been as crisp as people wanted yesterday. The Q&A got a little bit confusing. So just to be super clear, FY '23, last quarter, we had indicated that we expected 30% growth in the business, then you layer on the contribution from Innovium on top of that. If you look at our -- and so that -- so we still feel pretty good about that. And what I actually said was because our Q1 was a raise versus where consensus was that basically we're pleased to be ahead of schedule, right? And so if anything, it's going to be tracking higher than we expected. And so that -- just to be clear on that, FY '23 is on track or better than we had anticipated. And then kind of the other commentary would be, just given the tremendous strength in bookings we saw even last quarter, in our fourth quarter, gives us very high confidence in that outlook, plus the line of sight we have now to the new product ramps, particularly in the cloud that are going to contribute meaningful revenue in fiscal '24 and '25, that's sort of a $400 million of new business going to $800 million. That is now tracking to where that's going to -- we can see when those products are going to start ramping at the end of FY '23. And so yes, so strong backlog, strong order position. And then, of course, we have other design win momentum, which we can get into across the business, which -- more layers in '24, '25, '26, but that's just sort of -- so if anything, to be honest, everything got better versus where we were a quarter ago in terms of the outlook across pretty much the entire company's set of product lines and end markets.

Sure. Yes, I know there's just so much near-term results and people are focused there, but we wrote about this in our note that the focus really should be on the out years here and how you have these opportunities that are in layering on top of one another. And that's really what I want to spend the next 40 minutes talking about are those opportunities and some new ones that are emerging. And yesterday was one of my favorite calls that you guys have done because you, I think, scratched the surface on some of these new opportunities. I guess, first of all, this is a thesis I had a long time ago, and that was that, in some ways, you were rebuilding the Broadcom of old, the Broadcom before Hawk, maybe with the 2 Henrys back in the day, and you were focused on storage, network and compute and building those products that basically are the legos that you have to build custom products across all of your end markets really. I guess, first of all, would you say that's a fair assessment? Are you there? And yes, let's talk about these capabilities.

Yes, it's a great question, Chris. And I think, first of all, it would be always an honor to be compared to the original Broadcom. The 2 Henry's founded something very special. They're -- it's a little different in the sense that I think they had a very broad view, right? We're basically going to control all the hops within the network, the hops from the devices into the network, between devices. It didn't matter, right? And so you ended up with a portfolio of, call it, Broadcom classic, which had everything from Bluetooth to GPS, to WiFi, to switches. So crossed a myriad of end markets from consumer through computers, PCs, all the way into the network. I think our really unique thing that we sort of carved out, pioneered, if you will, was the concept of being the unique data infrastructure company, which in theory would be you can view it as a subset of what was being done, but actually, that's turned out to be where the puck is going. So we've been explicitly sort of in all the things that we had that were kind of like that. We could have rebuilt it, right? We had G.hn product line. We had a multimedia processor product line shipping into...

WiFi.

And things like WiFi. So we had pieces of that, we intentionally deemphasized and carve that out. And so I think we're building something that's unique and differentiated and hasn't been seen. And I think that's kind of the pitch I'd make to the investors. If you want to really bet on that trend, in particular, the build-out of the global data and digital infrastructure, being potentially as important as the physical infrastructure of the planet, we're the pure play. You can bet on to go do that. And so then as we talk about opportunities, they all flow into that. And I think that power of focus, Chris, has actually been a game changer for us in terms of focused set of customers, focused set of end markets and really being very purposeful that all the product lines and all the technologies we develop actually all hang together. And so we get a lot of reuse as you think about building platform technologies. We can leverage these across these sets of end markets that we focus on.

Yes. And I want to hit some of these. Let's start with SSD controllers. So first, we can talk about Amazon, they've been public now about that custom SSD controller, which I think is super interesting. You announced 1 SSD controller with an OEM. I think 2 more you might have announced last night if I got that correct as well. So let's talk about this. Is this a kind of great beachhead for you? Is this the beginning of something bigger? And yes, anything else you want to add to this? I mean you have a consumer, console, custom, semi-custom part as well. I mean this business is now starting to take off for you.

Yes. I would actually frame it more broadly, Chris, if you just talk about this as a category of really the focus we put into building a leading portfolio of data center storage, right, which is, as you say, there are controllers that we offer as merchant as well as we do customization for people. There is storage acceleration, right, which is a new market we highlighted, Dan highlighted at the Investor Day, right, which is $0.5 billion kind of emerging market. We've had great traction on the near line and sort of cloud hard-drive-based systems, which rows to the growing market. We've attached preamps to that. So I would really talk -- when we think about our storage strategy, it's not just SSD and kind of we made a controller here or there. It's actually the combination of having all of those capabilities makes us really a thought leader within that. Just to highlight a couple of successes, we've really driven the road map hard right there. So we said on the call, look, we've won multiple designs in data center-class SSD platforms with our PCIe Gen 5 solution, right? And this -- historically, there was a period of time when NAND companies were kind of all trying to do everything themselves. We sort of had a view that said probably the client goes that way. And then for the higher-end things, it's probably a mix. And if we execute right, we should just make it easy for them to decide to go with us, right, versus sort of burning their own cycles internally. And we've had a lot of success on Gen 5, right? We just announced we had another NAND OEM give us a design award. So that will be very successful. And then the Gen 6 product, which is in 5-nanometer, having that road map, I mean, that puts you 2 or 3 generations ahead, right, just on process node of our standard competitors plus anybody that's doing something merchant, nobody's going to go -- I mean, internal, going to 5-nanometer is just -- it's a huge lift, right? So I actually feel really good about that side of it. That's the focus of the group. We do have some other interesting things, as you mentioned, the semi-custom product that goes into game consoles. That's driving a lot of revenue. We've got a long-term road map there and partnership. And that's an offshoot, if you think about it, Chris, and kind of a leverage that we get off of the core investment. And we're willing to go do those things even if it's in the consumer market, if it's in something that's got some growth and some longevity to it. So I kind of wrap all of storage in that, but that has been a home run business for us that was basically left for dead 5 years ago. It was almost like -- it was like roadkill. Now it's a booming business, growing, strong operating financial performance and very differentiated.

Something that was missed by, I think, most in the Amazon announcement was the Nitro storage networking cards as well. Should I think about this as like a DPU but an SPU, so to speak, like a storage processing unit? And are we just scratching the surface there? Or do you think this is we're trying it, we're seeing how it's going to work? Or do you think other hyperscalers are going to get on board here? And then also, do you think other hyperscalers are going to get on board with the custom SSD controller as well?

Yes. Well, I think AWS is always a thought leader, right? And they do a great job and invent every year. And those trends that they highlight typically end up being indicative of where the market is going. In fact, a lot of the other companies wait to see sort of where they're heading, right? So I think those are definitely trends we see in the convergence of storage, security and networking converging in some of these different applications for offload, whether it's sitting on a net card or actually in a different part of the system. So I think those -- that can -- and having all of those IPs, if you think about a Marvell, is actually critical, right? Having CPU, having deep packet inspection and security, having the networking and then having all the proven -- Flash interface is required, it's a unique combination that we can deliver on. And so that -- I think the trends you see from AWS, I'd say, broadly speaking, are -- they've typically been the thought leader there.

Chris, we did say last quarter, like storage acceleration is clearly part of our overall cloud optimized platform. Think of it as another DPU offload essentially. So just likely the design we want to go across things like AI, ML, video, right? Storage is now -- security and now storage is one more. And storage, in particular, is a very important pretty data-intensive tasks in the data center, prior to end up using it for driving other value-added services. So, no, I would agree with you that, yes, I think it's a tip of the iceberg that you'll see a lot more of it going forward. In fact, we even size the market saying, "Hey, this is a brand new market. We did at Investor Day, I can put the slide up. It's a pretty nice opportunity for us.

Yes, it's just -- it's a nice intersection of networking and storage. A lot of guys have storage, a lot of guys have networking, but there aren't that many people that have both.

That's right.

So there are a lot of -- there are more people in the DPU space now, but this kind of SPU space or whatever you want to call it, there may be fewer players.

Yes. On top of that, I think the ability, as Matt said, to then also optimize it to any particular -- remember, each cloud is different. Within each cloud OEM, there's multiple clouds, right? And they all have their own different needs and different characteristics. And our ability to tailor solutions to each one of them, I think, it's again another big differentiation. It's not on just the IP part, which is, yes, you need to have networking and storage and security and compute and be able to put all that together, but the ability to tailor the drive to the customer's application, I think, is again another big differentiator. And that's been part of the reason why we'd be filing up these design wins, which are now going into production. Some of them as early as this year, right? Some of these going to production later this year.

Yes. Switching gears a little bit here. Another big takeaway from the call yesterday was this custom enterprise networking product or opportunity. So I'm familiar with Merchant. Just to catch me up basically on where you are there as far as I was concerned, this was kind of 10G campus on the enterprise side. And maybe Innovium can work into that, too. I was unfamiliar with this custom side.

Well, maybe -- Yes, I think, Matt, this is part of a much bigger narrative on enterprise. This is one of them, but maybe you can start off by -- because I think enterprise is actually going through a much bigger transformation. This is almost like one additional proof point. But yes, Matt, if you can start at a high level and [indiscernible] talk about this.

Yes, let's team up on this one. Let me frame a little bit of how that fits in, Chris, into the broader trends we're seeing in enterprise, and then I'll give you a perspective on the engagements. It's not just a one product, we got one thing and now it's -- there's just -- it's a whole new kind of product line for us, if you will, or revenue contributor. And then I'll let Ashish comment on some of the types of applications that we see. But it's not a one-off thing. And I'd say, look, if you go back more broadly, right, and you're seeing this in Broadcom came out with great numbers as well. And it's -- and you can see the OEM commentary from their earnings calls, right? Enterprise is extremely strong right now. And then within that, there's a couple of things happening, right? One is we've been talking for several years about our own efforts to win incremental platforms we didn't have before. We've gained content through those and kind of with our refreshed Ethernet portfolio. Those products now, Chris, as you see the OEMs revenues ramping, all those new products have Marvell content in them now, right? And that content, they're either new sockets or as you go from 1 gig to multi-gig or you can go through all the different things, there's a content gain that's happening. So you got enterprise as a market doing well. You've got these new platforms coming out that we're a part of. And then our content kind of per ship, per port has also gone up. On top of that, we've added -- with Cavium, even we added CPUs, right, as a contributor of -- that was one of Cavium's actually long-standing biggest end markets, right, was actually selling ARM-based CPUs or MIPS before that into these applications. And then what we highlighted is we've also now got -- and we had security also from Cavium, and now we have custom silicon multiple programs that are starting to ramp up with multiple companies. Again, leveraging our strong data-centric IP portfolio in latest process node with the road map to go to latest process node. So yes, we have -- and what's interesting is in some of these ASICs for example, that we do, we actually pull through other Marvell content with it. So it all hangs together, if you know what I mean. And that's why I think a lot of investors are scratching their heads. How is this market able to grow at this rate? Well, those are some of the reasons. And quite frankly, I mean, Ashish and I were joking about this, I'll just sort of have to admit it. I mean we undercalled it, Chris. And I think if you look back to the last 2 Investor Days, I think at some point we had to really focus the message right and the message was this focus of going from kind of being enterprise to 5G to cloud with an auto kicker. And enterprise has been a great performer all along. It's been a solid performer. And it's not like we didn't see this coming, but I think -- now that it's here, we're trying to explain why. And the exciting thing about it is it's not going away. We see multiple years of growth in front of it. So while it's not as exciting as some of these other end markets, it's a solid contributor, okay? And you have that enterprise base, and it gives you more scale. It has all kinds of benefits, right, that come with it. And it doesn't take away from the fact that our cloud business is doubling year-over-year. Our 5G business grew 30% last quarter. It's just another leg to the stool. Ashish, I don't know if you want to give one more click down on kind of how you're going to characterize the custom side, but it's part of a much bigger thing, Chris, that now is a business that's in the -- whatever it is, close to $300 million a quarter kind of revenue rate.

Yes. And Chris, I mean, these are typical applications, right? So this is typically -- basically access switches, right, during your first level of aggregation. But I mean the demand on these switches is significantly different. We found out ourselves when we went back into work starting a few months back. And we're also used to doing Zoom, and we were trying to do the Zoom from our office building and suddenly realizing when there's 100 of us going on Zoom, we literally crashed the network. Our IT guys are running around trying to get things fixed up, had security issues because you've got half the people in the building, half outside. So I think you're in the midst of this big upgrade cycle, which I think is really a transformation. And, look, if you talk to Jean, our CFO, right, she's not going to let this -- she can't fund this quickly enough to do it all in 1 year, right? And even really we wanted to, there's not a product out there. So I think what you're seeing is the start of what looks like to be a multiyear transformation. And I think what we are seeing is -- I mean, we knew about these ASICs in the past, right? These customs we'd won them. But the demand for them, right, has taken off and especially as this focus on new products has come out, right? We're seeing higher attach rate of our DPUs, right? Because you've got more compute functionality required, right? There's more intelligence on that, there's more end-to-end visibility that admins need. So there's a lot of content drivers really kicking in this business.

Great. Yes, I thought that, that was an interesting takeaway. Another interesting takeaway is around optical. There's a bunch of stuff that we can talk about there. 800 gig, you guys now have a product. You were already leading in 400. This is on the PAM4 side. And then ZR, 400ZR and ZR+. This has been delayed. People thought this was going to ramp 2 years ago. Now it's finally upon us. I don't know if you want to talk about these opportunities, and do we now accelerate on the optical side for you guys?

Yes. Is, do you want to lead us off on this one and I'll...

Yes, absolutely. So first, I think, Chris, is, I mean, the Inphi acquisition, I mean, talk about a home run, right? I mean this has been absolutely amazing. I think what they've been able to do for the company and really accelerating our cloud vision is just massive. A lot of the cloud optimized wins, quite frankly, they've been instrumental in the relationships, right, and our ability to accelerate to get those things. As you said, right, in PAM4 200, 400 gig absolute market leader, that continues. We announced the 800-gig product, which is PAM4 for the next generation last year, and now it's in volume production. It's not that the product is available, it's in volume production. There's optical modules in volume being produced by our partners as we speak. So I think that's kind of the big thing, and that's going to continue driving the growth. And I think as a lot of people know, right, there's only been a few hyperscalers, which have actually gone beating 200, 400.

Yes.

There's much more [indiscernible] market, very early -- like very early days, right?

It's like one major and then half is the way to think about it. So how do we think about broadening there? Is it 800? Is that the point?

No, no. I think what you're going to see is both, right? You're going to see both. The first is the folks who hadn't moved from NRZ to lower speeds, right, those are transitioning to 200, 400. So that cycle is going to continue. But then some of the more advanced adopters I'm looking at 800 gig as kind of the next upgrade cycle. So both things are going to happen simultaneously. Total volume is going to grow, right? There's more shift from the older NRZ standards to PAM. We'll benefit in both 200, 400, but we'll be able to drive the most premium product in the market, which is 800 , right? So that business is going to do very well. And a lot of people forget that it's not just about the PAM DSP, by the way. We provide the entire platform. So if you're an optical module partner, right, you're looking at us, you get the DSP, you get the laser drive, and you get the TIA, right? And you get -- we also announced our silicon photonics platform, which is a very integrated, [indiscernible] broad based modulators, prebuilt in, right. It just simplifies how you assemble the product. So our success in this market is not just a single product. It's really the whole platform of what you take, right? So that's what's happening on the PAM side. On ZR, yes, I mean, look, we pioneered the market, right, working with a key partner at 100, which was kind of a proprietary -- we were the only source. So we recognize that this data set interconnect market can be a lot bigger, but you need to make it a market standard. You need other players, quite frankly, right, to get involved, and that's exactly what we've done, right? And the 400ZR product is now shipping in volume. In fact, in it's very early days in the 400ZR market. But the start of those volume shipments, the total revenue on a quarterly basis in our DCI product line, in it's very early 400 is already higher than the peak we hit in 100. And we're at the very beginning of the cycle, right? So this is going to be a significantly larger market, not just within cloud customers, but also within the telecom market, right? And then maybe you can touch upon it, and Matt can talk a little bit, but we have another -- we did a new product entry as well, right?

You're stealing my thunder, Ashish.

Yes. Let's pause there for effect. Just to add, I think that was on -- I would just add a couple of points. I would say one is to reemphasize the Inphi transaction, I think you could argue is -- well, that and Cavium, but for different reasons, really, the most consequential or transformational thing that we did. The Inphi one, if you would recall, I know at the beginning, some of the multiples were a little bit eye-popping. But the reality is we knew, right, through the diligence, and I have confidence in Ford and his team very. I've known Ford a long time, very credible group. We obviously knew there was some upside we could go drive. And if you look at it, it turned out to be accretive, first quarter out, right, when we announced it, they've blown through the deal model, okay, for calendar '21. And you should assume they're tracking well ahead of the deal model that we anticipated for calendar '22. So those multiples actually look very reasonable now. If you just sort of throw them up against kind of what we pay and what the business is doing, you wouldn't even sort of blink now. It would be okay, that kind of makes sense. And then -- so we feel good about that. But I think the broader one is the kind of the customer revenue synergy. It is revenue's synergy because we don't sound like you can pull -- but the pull-through on design wins, the pull-through on the relationship with these customers. And then that enabled us to do Innovium -- And now Chris, we're sitting here planning our switch road map. And in the same room, we've got our optical DSP guys planning their DSP road map. And then as we get to the AEC market, we actually know what that looks like. And then we also have our NIC strategy. So in some cases, we're doing custom NICs, right, with interfaces that can. So we -- I think we have almost unparalleled insight into how these different architectures are developing. And then on top of that, each of these solutions is more than just 1 chip, like in the case of the optical module market, it's a chipset. And I refuse to go after this. I mean I have guys in the Marvell telling me, "Hey, we can go build up PAM4 400-gig DSP." Okay. Where's your laser driver? Where's your TIA? Where is your module relationship? How do you -- so when you hear about like, hey, I've got a part -- this is not the analog market. "Oh, I have a data converter -- he has a data converter, let's go call our data converter, who has the best SNR -- so it's a total system solution plus the link back to all the pieces kind of hanging together. So Inphi, at the time, you'll say, why did you buy an optical company. It's not a -- it's a high-speed networking capability that enabled us to do things like Innovium as well and get into this AEC market now, that's where you wanted to go. But I'll pause there -- there's a lot of adjacencies now we're able to drive, right? A lot of adjacencies.

They do strategically work together very well at this point. I'm not going to get into it on this call, but onboard optics and then eventually silicon photonics as well, and that all works into this as well. But let's now talk about active electrical cables. Something that probably just a few months ago I had never even heard of but now is a thing. And I thought, Matt, you made a great point last night. This is a potentially billion dollar market. And with the DSP technology that you guys have from Inphi, it's in some ways, transferable. So talk about this. Credo, people now know as they've entered the public market or getting a little bit more familiar, they talk about pioneering this market. But maybe talk about your position, how you're going to play here and -- and what kind of revenue and when.

Sure. Well, just a little -- yes, a couple of comments. I think it is very much a real market. I think the fact that I think Credo was able to successfully IPO and they've done a great job through that process. And they've done a great job educating the industry, right, about this opportunity. By the way, they're a fellow Marvell alumni, right, that are running the place, and I hope they will do really well. And I think they've done a ton of work, right, because they're laser focused in this area. So they've laid it out pretty, pretty effectively in my view. And like all emerging nascent markets, there's going to be a lot of opportunity for everybody, right? So this isn't, "Oh, well, we're going to go when and then they're not going to have any." I think both companies have a different approach, right? They've gone totally vertical, and they have a strategy there, right, working with customers to make everything from the cable to the electronics inside. We're a little bit broader, right, and that we actually work with a broader ecosystem of partners to go bring this to market. We're very confident in our capability right, relative to just having proven PAM4 DSP technology in high-volume production. We know well through our module experience how to work with the industry, right, to drive ecosystems. And we also have a very tight and deep relationships with the cloud companies. So I think it's going to be an exciting kind of new market that's emerging. And it's one that we're going to be very active in. Ashish, I don't know if you want to add anything?

Yes. No. I think just from a timing, I mean, our product is already available, right? So we have already -- this is basically reusing our existing PAM4 DSP, which is proven in high volume. So we basically announced the availability of our products. Those have already been shipping, right into our module partners who have already completed their AEC cables. So there's products available from our partners, right? And you can imagine they're going through the typical working with multiple cloud customers. So this is happening. It's happening actually very quick. This is not a typical you win a design, it takes 1.5 years, 2 years to get revenue. This is better than the quicker. And it's all because there's a transition from 50 gig per lane to 100 per lane. That's what's driving this essentially the data center. So it's happening as we speak.

Yes. Got it. Got it. I wanted to finish in terms of these new opportunities with carrier. You talked about a new radio win last night, which I think was interesting. But then also, if you could just update us, there are some other things here as well. So we talked about these Open RAN accelerator cards with Dell. Previously, you've talked about a massive MIMO opportunity with [indiscernible]. I think maybe there was some sort of ARM collaboration there as well. Maybe talk about these new opportunities? And then also, is this market ramping the way that you thought it would? And we haven't heard in terms of your 2 or 3 major customers. We haven't heard about new content layering on top. So is it just about units now of execution? Or is there still a lot of wood to chop in terms of additional content at those OEMs?

Yes. No, I think I kind of bucket your comments in 2 pieces, right? One is sort of the wins and the activity. And I'd just say broadly, we're in the middle of this market, right? I mean I think we hit it from all sides, right, from custom ASICs to Fusion basebands from CPU, DPU, switches, fronthaul chips, massive MIMO. So we're very active here and so some of the things you see are examples of that. Collaborations with ARM, great partnership with Dell in their effort to expand into this market. Those are just some examples. We did have a new incremental win and in the radio head on a 5-nanometer ASIC that was more to just show them the momentum continues. I'd say the biggest development of which we feel really excited about is that we made a bet on DRAM a few years ago and carved out a separate product for that and separate sort of go-to-market strategy. And that's gone well. I mean, Dell is one example, but we've highlighted previously at least 3 other important wins, right, that ultimately inflect into a hyperscale environment. So I think I'm glad we made that decision, right? That was incremental R&D. It was a little bit of do we go chase this because there's always been open standard concepts in the telecom world that end up getting killed, but this one feels like it -- not only is happening, but it could be incremental because a lot of the use cases now are outside of your traditional sort of telecom model. And they're really -- the trend has really been around private enterprise 5G. And I've heard that not only from the system base station OEMs we do business with, but also from the cloud side. So I think that could end up being, Chris. We don't know for sure yet, but it very much appears that DRAM and enabling that kind of layers an incremental opportunity versus it's just cannibalistic and it's just removes it from a base station shipment somewhere. So I think that's all tracking well, to be honest with the investment community. We don't get a ton of questions on 5G anymore. It's ramping like crazy. It was up 30% sequentially in our fourth quarter. So it's kind of what we thought was going to happen. If you -- even if you go way back, right, at some point, we had given a marker of like $600 million kind of revenue we could achieve, and then we bought Avera and then we won more business. And you can kind of take our Carrier number and you sort of know those ratios. We're already sort of tracking into where we -- even the base case where we were. There's more to come. And we haven't even had -- we got a whole ramp year ahead of us in years in '23 and '24 just on our baseband products, and you've got layer 2 stuff coming in probably later. So there's -- and then you got -- whatever happens with DRAM is a little bit of a call option. So I would just say, broadly, we're very happy with 5G. It didn't -- these things never happen exactly the way you think they will in terms of which customers ramp when and all the timing. But the aggregate is, because we're so broad here, Chris, in terms of our exposure on different OEMs, different kind of product cycles, the end result is it's growing very strongly. I think our -- Ashish, our carrier revenue was up 45% year-over-year. Is that...

Yes, it's a pretty big number. Yes. And obviously...

It's not your typical. And you know the reasons, right? We didn't have a lot of share in 4G. We've kind of hit it in 5G, content is higher. It's kind of the story of all -- a lot of our markets, if you think about it, Chris, right? We never have a big position in the last cycle. We've got a strong position in this cycle and contents moved up and the demand strong, right? Just business is good.

Design win question, like, I mean the reality is most of our wins, we talked about are already multigenerational. It's not only -- we have what we are shipping today. We actually have already locked up because that's how the nature of these infrastructure programs is, right? It's not like a -- you change every cycle. Once you've got a partner, you work very closely with, there's IP from both sides. It's a very collaborative effort. So not only are we shipping what we want for the first generation that's in production today, but we've already won right, the next generation at multiple customers. So this is a long visibility, long life cycle products. Yes. So this thing is doing amazingly well.

Great. And in the final minute, you can talk about anything you want, but I'm getting pinged with a lot of questions about supply. What you're ultimately going to do here? I got a question, would you use Intel, for example. What are you doing to solve this? And when can we see your lead times come down?

Yes, we [ will set ] the Intel one aside because that's way far away for anybody, right, who actually wants to engage in IFS. That's a long-term -- that's another discussion, right? Look, I think the way I would think about this, and I think this is a testament to all the work we've done over the last year, right, to really reposition Marvell kind of strategically in the minds of our supply base, if you think about it. How are we getting more supply right now, right? Because I mean, everyone is struggling and there's still a gap and -- but we're pretty happy, right? If you just look at our kind of year-over-year growth numbers we're delivering. That means we're getting more supply, right? Like we grew 37% or whatever, about high 30s percent organically, right, year-over-year if you sort of backed into Inphi plus Marvell. That means we got that much more supply. I think what's happening is those key suppliers are betting on Marvell, right? At the end of the day, right, everybody is worried about an oversupply situation. All the way down into the food chain, pretty deep, right? You get to the suppliers, suppliers and everybody is worried about a correction. So I would say the supply chain partners we have really believe our story. We continue to get incremental supply. I think the thing that's been unanticipated is that demand continues to be very strong. This is a cycle like I've not seen. And I think, on top of that, because we've hit the sweet spot on some of these secular growth opportunities, the pressure is even higher on us. But look, I think we're going to continue to just chip away at it, Chris. I think that's all you can do with. And I think, to be honest, to the extent that there is a slowdown in some of these other consumer-type end markets, you could say PCs or phones or -- it doesn't really matter. And I think there is a lot of a real high degree of investor concern on some of those end markets. We don't participate in those, so I don't have the insight into what's going on, but I sort of read the reports and can triangulate a few things. If there is a slowdown, that capacity -- we're sort of waiting at the front of the line here for it to move over. And we have seen pockets here and there. And that's part of that too is being kind of tactical about, "Hey, something opened up. We can get some x amount of substrate." I know you asked us a year ago, we didn't have anything, now we have some. So we're seeing that kind of a thing. And I think to the extent that another end market actually decelerates, I think that's going to be actually good for the industry and good for Marvell. But right now, everything is pretty hot still is my read when I talk to the supply chain partners. But our model all along, Chris, right, we're never -- we're not a consumer company, right? We're not somebody that rides a smartphone cycle and one quarter we just blow it out and then we have a big reset the next quarter. And if you invest in Marvell, right, you really are investing in kind of a long-term ideally, right? It's not going to be perfect, but more repeatable, monotonic kind of growth driven by these different end markets. I've worked really hard to diversify the company, right, in product lines and end markets we're exposed to so that you can deal with periods of digestion or lumpiness. I think we've been able to do a pretty good job at that. Even 5G would be a great example, right? We basically powered through the last 2.5 years. I think the worst thing that happened is one quarter we didn't grow, like it went down. It just like it was flat or it was like -- and then the next quarter it kept going. Normally, that kind of a market is totally lumpy. So I think that's what you should expect from us, continued sort of progress through the year. And obviously, as I've said, I'll close it out by saying we're very enthusiastic about our fiscal '23. We're ahead of track based on our Q1 guide. And we're layering in new content opportunity sockets to fuel the growth in '24, '25 and beyond.

Yes. I've never been more bullish on Marvell. And to your point, you made some [ bull ] calls with some of these acquisitions, and they're not only paying off now, but we're in the early innings of the payoff, I think. And so with that, I guess it's a great point to close on. Thank you, Ashish. Thank you, Matt. Thank you for your time. Really appreciate it.

Yes, thanks, everyone, today. Take care, guys. Bye.

Cheers.