Please welcome Marvell Chairman and CEO, Matt Murphy.
It's great to be here to kick off day one at COMPUTEX, and it's great to be back here in Taiwan. The first time I came here was nearly 30 years ago. It was my first business trip to Asia, and I remember back then visiting some of the key technology companies here at the time. Many of them were still young, small companies, emerging companies, and today, those same companies have become the most important technology leaders in the world. Now, I've had the opportunity to come back many times and see Taiwan continue to grow in importance as one of the world's leading technology centers. Today, so much of the future of AI infrastructure is being built right here. I have a question for all of you. What defines the performance of AI infrastructure?
Now, maybe you're thinking about the processor, the GPU, the XPU, or maybe it's the process node used to build it. 3 nm , 2 nm , or soon A14, A16. Those are great metrics. They tell you a lot about the speed, the efficiency, and the density of the compute. AI workloads are certainly compute-intensive, but that's not the whole story. Now you might say, "Well, what about memory?" AI workloads are incredibly memory intensive as well. More memory, higher bandwidth, all of that matters. It's all critical, no doubt. That's still not the defining characteristic of the system. Because one processor, no matter how fast it is, no matter how much memory it has attached to it, is simply not enough for today's AI workloads. You need tens of thousands and eventually millions of processors working together as a single massive compute engine.
That's why computing at this scale is fundamentally a connectivity challenge. Increasingly, it is the architecture and characteristics of connectivity that defines the performance of the system. Now look, we've seen incredible breakthroughs in accelerated computing, and we've seen the emergence of high-bandwidth memory to meet the AI challenge. I'm here to tell you the next major wave of innovation and scale will come from the underlying connectivity of these systems. As those connections move from copper to optical, they will unlock new architectural possibilities. Today I'm going to explain why connectivity is becoming one of the defining characteristics and challenges of the AI era, and why this technology transition matters to optics. Now, this isn't something far out in the future. It's happening right now, this year, next year. We're in the ramp.
At Marvell, we've been preparing for this moment for nearly a decade. We built the company very deliberately around the infrastructure required to move data at massive scale. To understand why we made that bet, let's go back in time, 10 years ago, when I joined Marvell as the CEO. Prior to Marvell, I spent 22 years at one company, Maxim Integrated Products, which was a leading analog semiconductor company. One of the unique things about working at an analog company is that your products go into virtually every piece of end equipment, every electronic system, every end market in the planet. Over those two decades, I had a front row seat to just about every major technology trend. First personal computing, then notebooks, digital still cameras, smartphones, eventually data center. I watched wave after wave of technology reshape the whole industry.
I joined Marvell, and I didn't start off actually thinking about, well, what products do we have? I reflected on where the industry was headed. It seemed clear to me, even at that time back in 2016, that the next major growth cycle for semiconductors in the world really was going to be driven by the data platform companies. Back then, it was still the same ones as today, companies like Google, Amazon, Microsoft, Meta, and more specifically, the semiconductor technologies that were required for those markets to move data, store data, process data, and secure data, do it at massive scale. That was the vision we had. When I looked at the products we had at that time, very few of these were actually exposed to that trend. It was kind of a problem.
Less than 10% of our revenue 10 years ago was coming from data center. That's it, couple hundred million bucks . More than 60% of our revenue back then was coming from consumer, it was exciting time. We were in virtual reality headsets. We were in gaming consoles, streaming devices, wearables. In fact, our claim to fame back then was Marvell was designed into the first Wi-Fi connected Barbie Dreamhouse. That was our big design win. It was real. In fact, the first week I was at Marvell, the team briefed me on what a great design win this was. That's where we were. We had a vision. There was a pretty big gap, though, between the reality that we were facing and where we saw the industry heading. We had conviction. We decided to bet the whole future of Marvell on it.
To do that, we needed a clear vision, and our vision at that time was pretty simple. By the way, this is still the same vision that we have today, 10 years later. Which is build a best-in-class pure-play company focused on semiconductor solutions for data infrastructure. At that time, data infrastructure was not a recognized market category. It was the term that we used to describe the infrastructure that was going to be required to move the world's data, store the world's data, process the world's data, and secure it. Like I said, we were not in that business yet, and frankly, we didn't even have a lot to work with as we went after it. We had some.
My team and I came to a conclusion, which is that we would need to build these capabilities internally, and others we would need to build through strategic M&A. We had to get focused because when you're transforming, it's not just deciding about what you're going to do, it's equally important to decide what you are not going to do. With that strategy in place, we got to work. We began systematically building Marvell around that vision, and it wasn't just one move. There was a series of deliberate choices. We looked for the premium assets in the markets that mattered the most, the best companies, best technologies, the best teams with the strongest market positions. We first started by divesting businesses that weren't aligned with our strategy. You can see some of those there.
Very quickly, we acquired Cavium to strengthen our compute and networking capabilities. That was back in 2018. 2019, we divested our Wi-Fi business. Again, we were focusing. We acquired Avera to establish our custom silicon business and Aquantia to bolster our connectivity portfolio. In 2021, we followed all that up by acquiring Inphi for $10 billion. It was our largest acquisition to date, and we got world-class data center connectivity technology into the company through that. We acquired Innovium the same year, adding high-end data center switching capability to the portfolio. We took a break. We took a few years to digest and focused on unifying and building out our whole technology platform to address the data infrastructure opportunity. Over the last 12 months, we fired up the M&A engine again.
We divested our automotive Ethernet business, again, power of focus, and acquired Celestial AI for its Photonic Fabric technology and XConn for scale-up switching. If you add it all up, over the last decade, we've invested roughly $22.5 billion through acquisitions. We spent $18 billion organically inside of Marvell to develop the platform. We divested approximately $4.5 billion worth of assets. All in, we've invested roughly $36 billion investing in this platform. Now, let me show you the result of some of these investments. First of all, we have built an incredible technology platform. It all starts with the advanced process node. It's one of the most important decisions we made, actually, was to become a process node leader.
Marvell, Cavium, and some of the companies we acquired had all been fast followers, meaning you're like a node or two behind on everything you do, and that's largely a result of just not having enough scale. That's usually why people do that. As we integrated these businesses, we made the decision that if we're going to compete in data infrastructure, we had to be at the absolute leading edge. No choice. Here's a little-known fact. Marvell skipped 7 nm completely. We made a full node jump at that time from 14 nm and 16 nm all the way to 5 nm. I mean, nobody does this. Nobody takes that kind of a risk or a bet, but we did, and it worked. It worked really well. Flawlessly, actually. Our engineering team did an outstanding job executing this transformation.
In early 2020, we released our first world-class IP platform, complete with die-to-die interfaces, custom SRAM, high-speed SerDes, and more. SerDes is a good example of how we built this platform. It combined Marvell's own core engineering strength with exceptional talent from Avera, Aquantia, Inphi, and others. Today, that is a 1,500-person organization at Marvell, second to none in terms of engineering scale and capability. To support the process data portion of our mission, we built a best-in-class custom compute platform, working in deep partnerships with the world's leading hyperscalers, and that business has been doing very well for us. In store data, we built a whole portfolio of storage controllers, CXL-based memory poolers, and near memory compute. Here's where we really went all in. That was in data movement.
This is where our high-speed connectivity portfolio, and when you look at Marvell's data center business today, the vast majority of our revenue actually comes from connectivity, from high-speed optical interconnect inside the data center to long-reach optics between data centers to high-speed switching infrastructure. Today, we are the undisputed connectivity leader. When you step back and look at what we built and where the market ultimately went, I think the results speak for themselves. Back in 2016, Marvell was a $2.3 billion company. As we embarked on the transformation, actually, in the first five years, we doubled the company, $4.5 billion in revenue. Over the next five years, our growth accelerated, and according to consensus estimates on Wall Street for the current year we're in, we're set to grow about 2.5x over the last five years to $11.4 billion.
In the recent couple of years, if you actually drill down, Marvell has been growing like 40% a year. The growth rate is actually accelerating in the last few years. At this point, Marvell is off to the races, okay. Based on the outlook that we shared in our earnings call last week, consensus estimates have come up and they expect us now to deliver $16.4 billion in revenue next year. As I said earlier, when we started this journey, data center represented less than 10% of our revenue, and we bet the farm on it. Last quarter, it was over 75% of our revenue and growing very rapidly. This is a very different company than we used to be, and the thesis is largely played out. We're still in the early innings of this infrastructure build-out.
The next phase is all in front of us. It'll have a different set of requirements, and that brings us back to connectivity. For the past several years, as AI has created new demands on the infrastructure, we've seen the industry solve one major bottleneck after another. First it was compute. The industry needed dramatically more compute to enable modern AI. NVIDIA did an incredible job leading that revolution, along the way became the world's first $5 trillion market cap company. Congratulations to Jensen and his whole team that's here. It's just a phenomenal result. Next came the memory bottleneck. Larger models required enormous amounts of memory and bandwidth. The memory companies are scaling aggressively now to meet that demand. Just recently, we've seen three new $1 trillion market cap companies emerge in that market. The bottleneck is shifting again.
It's connectivity that will define the limits of the infrastructure, just like with compute and memory. The industry will rally to meet this challenge. This isn't just me saying this. This is what we're hearing from our largest customers. The world's largest hyperscalers are now reimagining their entire network architectures. They recognize that scaling AI infrastructure is now, first and foremost, a connectivity challenge. As reasoning models, mixture of experts architectures, agentic AI, it all continues to evolve, more data has to move across the infrastructure, demanding higher bandwidth and lower latency. As workloads no longer fit within one data center, guess what? They need to build larger data centers or full campuses full of data centers and all the high-speed connectivity between them. The connectivity becomes a critical enabler of scaling compute.
Increasingly, our customers recognize that optics is the way forward, and they're looking to leaders like Marvell to help them build larger, faster networks and at scale. When you look across the semiconductor industry at the leading companies supporting this infrastructure build-out, it becomes clear each of us is focused on a different part of the infrastructure. That shows up in the revenue mix. Some of the companies are compute first. It means the vast majority of their revenue is tied to compute, with some of it tied to connectivity, but most of it's compute. It's obviously a critical part of the stack, and that's why we have several trillion-dollar-plus companies in this group. You have the companies focused on memory, and again, all trillion-dollar market cap companies at this point. It's unbelievable. NVIDIA and Marvell. We're different. We're unique.
Today, the vast majority of our revenue actually comes from connectivity. We built this company around data movement, and today the vast majority of our revenue comes actually from connectivity. This spans a broad range of technologies, and even the portion of our revenue that's from compute, which you can see, is fundamentally because customers embed our connectivity in their compute engines. This gives us a unique position and perspective on these technology transitions that are happening. It creates a very different relationship that we can have with the rest of the ecosystem. We partner deeply with the compute companies. We partner deeply with the memory companies. These are very strategic relationships, and in many ways, we are the Switzerland of the industry, and we work with everybody.
One of the best examples of the role that Marvell plays in this ecosystem is the recently announced strategic partnership and expansion with NVIDIA. As part of this announcement that we made a few months back, NVIDIA invested $2 billion into Marvell. We're expanding our partnership now across multiple dimensions, including optics, photonics, NVLink Fusion. I'm thrilled to announce that Jensen himself is here today. He's going to join me on stage. We're going to spend a few minutes chatting about the partnership, and we're going to see where AI infrastructure goes from here. With that, let me please welcome to the stage Jensen Huang.
Hey, man.
What's up, Jensen? How you doing?
Boy, that's a huge stage.
Good to s-
I had to run a long ways.
Are you out of breath? You okay?
Oof.
I know.
Scared me.
Let's fire up. Good to see you. There you go.
Nice to see you.
Yeah. Congrats on a great kickoff yesterday, GTC. You guys are off to the races this week.
Thank you.
Look, maybe you heard some of what I just said. We're talking about connectivity today.
The next trillion-dollar company, ladies and gentlemen.
Whoa. That would be exciting. Let's do it together.
Yeah.
It really all starts with what's happening today in AI infrastructure kind of more broadly. How do you see that, just from the big picture standpoint? We're at this extraordinary moment, customer demand's through the roof. How do you see connectivity playing into this in the interconnect that's required?
Yeah, that's really great. Yesterday, I said that useful AI has arrived. It's the reason why your demand is going through the roof. It's the reason why my demand's going through the roof.
Yeah.
This new computing pattern that makes it possible is called agents. These agents has a particular computing pattern that is disaggregated and distributed. When you take a computing problem and you disaggregate it into a lot of parts, and you distribute it across the entire data center, what's necessary is connectivity. That's the reason why Matt's doing so well. That's the reason why Marvell is so essential. We've distributed and disaggregated computing so that it runs across these enormous clusters so that we're aggregating the total compute, the total memory, the total-
Yeah
bandwidth that we have, and what makes it possible is connectivity.
Yeah. We're seeing it.
That's why they're going to be the next trillion-dollar company.
We got a little work to do, but we're on our way. We're on our way. Thank you, Jensen. Well, let's talk about scale. We used to talk about 10s of GPUs, and CPUs, and XPUs connected, now thousands, now maybe millions at some point. As you scale the compute and you scale the connectivity, I think we talked about things like agents, but how do you think about that across data centers, within data centers? How do you think about connectivity at large playing that role, and what kinds of technologies do you think are important there?
Well, at the foundation of it, the agent computing pattern requires an orchestration system that allows the large language models, the computing, to be able to think and reason and come up with plans. It also has to use tools and browse the internet, access memory, access long-term memory, deal with short-term working memory. All of that requires a lot of connectivity. It's also the case, and if you look at the way we introduced Vera Rubin, Hopper was designed for training.
Yeah.
Grace Blackwell introduced GB200 NVL72, our first scale-up fabric. It introduced the idea of extremely fast inference for MoE models that are very large, mixture of expert models that are extremely large. Grace Blackwell was for inference. Vera Rubin is to run agents.
Yeah.
Which is the reason why the Vera Rubin system includes, of course, the Vera Rubin thinking AI, but it also includes Vera CPU for orchestration. It includes Vera CX for storage acceleration, for managing long-term memory. The way that I think about these systems, sometimes maybe the CSP wants to design their own custom chip. Between us, we also partner together on NVLink Fusion.
Yeah.
Which makes it possible for you to use the same system architecture, and with Vera Rubin inside some of your semi-custom chips.
Yeah
a lot of your interconnect silicon photonics and optics and technology such, and we can create essentially a disaggregated, distributed, and heterogeneous data center. That's the big idea.
Yeah.
Yet their system architecture is identical. Their networking technology can leverage a lot of NVIDIA stack. The CPU could be Vera, yet it can leverage a lot of your stack. NVLink Fusion is about taking NVIDIA's technology and our platforms, Marvell's technologies. We fuse it. That's why it's called fusion.
Yeah. No, I think about the partnership, and we've been working together a long time. I think memorializing it with the investment, which we really appreciate, I think it's been huge for us. We're honored to have it.
Who doesn't love making money? It's nice to give.
It's done well since you invested.
Boy
when Jensen invests, just follow him.
I love getting rich.
Just follow him.
Give Matt all my money and just watch it make money.
That's what I'm doing every day.
I love that.
That's what I'm doing every day.
Yeah.
I think these things you talked about, which we brought to fruition, NVLink Fusion, working together on optics, I think the era of agents and your new platform now, I think it's ideally suited. NVLink Fusion, we had this idea years ago, right? I think it was a little ahead of its time, and now, and I wanted to see if you agree, when you think about your platform and then some of the custom networking and compute needs that our customers have, and the ability and the need to interoperate and work together, it seems like the time is now between Marvell and NVIDIA to really go enable our customers to have that flexibility that they're looking for and really use the era of agents to scale our platforms together.
Yeah. Ultimately, I do think that if you buy nothing but NVIDIA, it's okay. Okay? If you absolutely must design your own ASICs, we're still happy-
Right
having NVIDIA be inside that data center. You don't have to buy everything from us. Just buy something from us.
Right.
We're happy to support you and support the customer. I think that between the two of us, you have the benefit.
Yeah
of a general purpose, very high efficiency, a system that is very well-built, starting with, of course, Vera Rubin.
Yeah
that you want to extend, to specialize, you can do so as well, which is the reason why your customers and mine, NVIDIA's in AWS, Marvell's in AWS.
Yeah.
NVIDIA's in all of the clouds, and it's wonderful to see Marvell expand into all of these different clouds.
Yeah. Great. Thanks. Hey, one last one for you.
Just leave some business for me.
Look, we're your best salespeople right now. Well, you have great salespeople, but yeah.
I'm your best salesperson.
Working together. Final question for you.
Yeah.
A lot of my talk is about some of the transition, especially as you go to inside the rack from copper to optical.
Yeah.
It's obviously not going to be a one zero. There's time and there's different use cases. How do you see that playing out right now, the transition from copper to optics, and maybe how we can work together there too.
Well, we should use copper as much as we can for as long as we can.
Yeah.
Copper has its limits. Copper has its limits with bandwidth and also with distance. Ultimately, the right strategy is to scale up with copper as long as you can. After that, you scale up further with optics.
Yeah
You scale out with optics, and you scale across with optics. You use optics wherever you must. You use copper wherever you can. I think that that intersection is going to continue for a long time. Here's the bottom line is, in the next five, 10 years, we're going to use a ton of copper, and we're going to use tons and tons of optics.
Yeah.
These data centers are part of infrastructure now. The reason why I say that AI is now useful AI has arrived, is because now AI is profitable and tokens are profitable. When token production is profitable, everybody wants to make more tokens, which is the reason why Marvell's demand is so high.
Yeah.
Our demand is so high because everybody wants to produce more tokens because it's used all over the place by agents.
Yeah, absolutely. Well, I think you touched on a bunch of things I'm going to cover later. If you want to do the rest of my presentation, you can, if you want.
Yeah. ladies and gentlemen.
These beautiful slides.
Yeah, Matt, just sit right there.
You take it from here. All right, Jensen Huang.
All right, you guys.
Good to see you, brother. All right, take care.
Okay, you guys. Thank you.
Thank you, Jensen.
Buy Marvell.
Buy Marvell. All right. Outstanding. Super fun to have Jensen here, as always. All right, we've been talking a lot about connectivity. Jensen and I just covered this. Let's dive in now. Let's go one level deeper. AI infrastructure spans every distance. It spans from hundreds or even 1,000 km between data centers to just millimeters inside the package. Every one of those distances, it requires a different solution. It's a different technology, different engineering team. It's a completely different set of experts, and in many cases, it's a different supply chain. These are not variations of the same problem. What you have here is fundamentally different engineering challenges, and that's what we're going to walk through next. All right, let's start with the longest distance. Jensen referred to this. This is scale across, connecting data centers together.
Every major cloud provider has hundreds of data centers around the world. All of those data centers need to communicate with each other. This is fundamentally a long-distance connectivity problem. We're talking about links that can span hundreds or even a thousand kilometers. This requires very specific, very complex technology called coherent modulation. At the heart of it is a specialized digital signal processor, or DSP. It's designed to push enormous amounts of data across fiber optic cables over very long distances with extremely high reliability. There's only a few companies in the world that build these coherent DSPs, and we're one of them. Marvell has been a leader in this technology for many generations. We build optical modules that contain all the electronics needed to drive and modulate the laser and transmit data over long distances. I've got a little show-and-tell here in my pocket.
I'm not holding up a chip this time. I'm holding up an optical module. This is one of our coherent optical modules. This is an incredibly complex piece of engineering. At Marvell, we build the entire module. This is ours. It includes the advanced node CMOS DSP. It's among the most complex chips, just the DSP alone, that we design at Marvell. It also incorporates inside our fourth-generation silicon photonics technology. That's inside here. We've been developing that technology and in production for a decade on silicon photonics. It also includes our own broadband analog components that we designed, which is designed in silicon germanium. Marvell pioneered this technology starting with 100 Gb per second a decade ago, then moving to 400 gig, and now shipping 800 gig in volume.
Later this year, we'll be sampling the world's first 1.6 Tb, 2 nm coherent optical solution. That couldn't come at a better time. Demand for bandwidth has never been greater. All right, now let's go inside the data center. These data centers can be very large, spanning hundreds of meters, and they contain racks and racks of compute servers. Each rack typically has a switch at the top with servers connected into that switch. Those rack level switches connect to the spine and then the core switches. This creates the network fabric that ties the entire data center together. All of that is connected through fiber optic cables. Once again, optical modules drive data transmission over those fiber optic cables. This time, the modulation scheme is different. Instead of coherent technology, we use a more power-optimized modulation technology, which is called PAM4.
The two key semiconductor solutions for this part of the market are the PAM4 chipset inside the module and the cloud switching infrastructure that ties the data center together. Marvell builds both. Starting with the PAM4 chipset, we build the industry's leading PAM4 DSP solution. Also the high-speed analog components that go around them, including transimpedance amplifiers, or TIAs, and laser drivers. These are also in silicon germanium, by the way. We've led the industry through every major transition of PAM technology, starting at 50 gig, 100 gig, 200 gig, 400 gig, and 800 gig. Last year, we began ramping Marvell's 1.6 Tb 3 nm PAM4 solutions, leading the industry's transition to 1.6 Tb connectivity. For Ethernet switching, Marvell has a similarly complete portfolio of products, from 12.8 Tb to 51.2 Tb.
Today, we announced our new 100T Ethernet switch, specifically designed for AI data centers with the industry's lowest power. Whoo. Special announcement for COMPUTEX. We waited. You put it all together, we provide a complete solution for connectivity inside the data center. Let's move inside the rack. The goal here is to connect the largest possible number of processors together in a full any-to-any configuration. In other words, every processor can communicate directly with every other processor. Jensen talked about this. The first company to bring this architecture to market was NVIDIA with NVL72, named for the 72 GPUs connected together inside a single rack. This required a completely different approach to connectivity. It was a different class of switch and the ability to drive very high-speed signals over copper backplanes inside the rack. Today, this is not the domain of optics.
This is the domain of copper. The core differentiator here is the electrical SerDes technology, not the optical. Marvell also has leading electrical SerDes at 200 Gb per second today. We've demonstrated already, over the last couple of years, 400 Gb per second for the future. We're building this SerDes technology into our customers' custom silicon and their XPUs, and also into our own scale-up switches. All right. Let's go all the way inside the package. Here, we're not talking about meters anymore. We're talking about millimeters. You might not actually think about this as a connectivity challenge, but today, most advanced chips have multiple chiplets inside the package. When you have 2.5D or 3D packaging, it's fundamentally a connectivity technology, actually.
It allows these chiplets to sit very close together inside a package and communicate through ultra-high speed, short-reach die-to-die interfaces. Marvell has leading die-to-die SerDes and leading capability in advanced packaging, allowing our customers to build some of the most complex, unique multi-die chips in the industry. As you can see, connectivity for AI data centers requires a very broad portfolio of technologies. Each distance requires a very different solution. Marvell has the industry's most complete portfolio, from millimeters to kilometers, every hop, every distance. It turns out having all of those capabilities under one roof is unusual. It's unique. When we go and compete, normally there's a different set of companies that we compete against in each one of these categories across these different distances. This is what makes us unique. We're the one-stop shop. We're the leader across the entire connectivity stack.
That brings us to the next major challenge facing the industry. What you probably noticed as I described these different solutions in the last couple of slides is there's different solutions for different distances, and that some of those connections today are optical, and some of those connections today are electrical. It's actually defined by distance. The connections on the left side of this chart are optical today. That means they use fiber optic cables to transmit light, with complex electronics on either side of the cable to drive and modulate the laser that's transmitting that light. The connections on the right side of this are electrical, so they use copper cables or just copper traces that are printed on the circuit board, or even microscopic copper routing inside the package. The common theme here is copper.
In the middle, you see the wall, the copper wall. The wall is defined by the longest distance you can transmit a signal over copper. Before you have to move to an optical connection. This is an important distinction because copper is simple and it's low cost, and as Jensen said, you want to use it for as long as you can. It's very practical. Optics, and optics is more complicated. It requires lasers, photonics, complex electronics. It's a bigger lift, but it's going to be needed. The copper wall, what I'm here to tell you today, is it's about to move. It's going to move again, and it's going to take over the rack itself. This is creating an explosion in demand for the optical industry. Incredibly complex engineering challenges are coming and along the way.
Why is this happening? It's not just somebody's preference to go do this. This is physics. The distance a signal can travel over a copper cable is inversely proportional to the bandwidth. Every time you double the bandwidth, you have to cut the distance in half. Today, the highest speed production systems in the world run at 200 Gb per second per lane, just to give you an example. At that bandwidth, the cable length is limited to roughly 2.5 m . By comparison, systems running at 100 gig could use about 5 m cables. The height of the rack is about 2 m . Once you account for all the routing inside the rack, 2.5 m is right at the limit. When we move to 400 gig, we can no longer fully connect the rack with copper.
The wall is moving, and it's moving now. Going forward, even the connections within the rack will become optical, and the whole industry knows this is coming. We've been preparing for this moment, not just Marvell, but the industry. You see this in Taiwan, by the way, in the supply chain and the ramp-up that's happening. The ramifications for this are actually enormous because each time the wall moves one step to the right, the number of connections that you have goes up by at least an order of magnitude. It's creating this explosion in demand, as I mentioned, and the optical supply chain needs to scale up massively and be ready. We've seen this movie before, okay. I mean, 20 years ago, and I remember this, when state-of-the-art was 10 Gb per second inside the data center. It was 10 gig.
We used copper cables all across the data center. Optics back then was reserved for just very long distances. It was essentially like a telecom technology. When the wall moved, the optics industry actually rose to the challenge. Today, all the hyperscale data centers in the world, they're all optically connected. As we saw in that transition, it did require new solutions. You couldn't use the same power-hungry kind of telecom approach, which is where PAM4 came in. It was optimized for power, density, and reach, and requirements specifically tuned to inside the data center. Marvell was one of the key innovators there. We're about to see the same wave of innovation needed as optics moves inside the rack. That's with a technology called co-packaged optics, or CPO. You hear a lot about this now. I'm going to tell you more.
CPO is a technology where we bring the optical connections all the way to the package itself, right next to the compute, either the custom compute or the switching silicon. The fundamental challenge we're solving with CPO is density and power. Now remember, the number of connections inside the rack is like 10x the number of connections between the racks. If you just try to use the same optical technology used across the racks in the data center, you wouldn't have enough power. You wouldn't have enough physical space. You cannot fit all these standard optical modules and cables as they are today. It just doesn't work. It's not possible.
The industry has been inventing this co-packaged optics concept, which brings the optical fiber right to the package, and it tightly couples the electronics that drive the signal over the fiber directly with the custom compute or switching silicon. This is a massive change, and it's hard because you're combining some of the most advanced technologies in the chip industry, leading-edge CMOS, silicon photonics, advanced packaging, optical interconnect, all manufactured in a small, tightly integrated system. The complexity is very high, but it's the only way to continue scaling bandwidth and overcome this limitation that I talked about with copper while reducing power at the same time.
This is where the industry's headed, and this is one of the reasons that Marvell has invested for more than a decade in silicon photonics, optical DSPs, all the analog broadband components around it, and all the advanced packaging you need to pull this off. It needs to all come together actually in CPO. This isn't some futuristic thing, guys, okay. It's happening now. In fact, I brought a couple of Marvell examples with me today. Let's do a quick show and tell. Okay. Over here, you have a traditional Ethernet switch. This is our 100T Teralynx switch that we announced today, and you guys are the first to see it, actually, everybody here in the room. You can see the switch in the middle of the board.
Copper traces inside the PCB carry the signal to the front panel, which is here, and this is where all the optical modules plug in. Let's move over here. This is a CPO-based switch right here. Notice that there's still the switch silicon in the middle. That's right in the center of the die of the package. In this case, this is our 51.2T switch, and all around the edges are 16 3.2T optical engines. The 16 times 3.2, you get 51.2. The fiber's directly attached now to these engines. It's not to the front panel. We've completely eliminated the copper traces on the PCB. Light comes directly out of the package. This is a very complex piece of engineering, and it was very cool to be able to show this off today.
Okay, co-packaged optics is here, the industry is scaling up to meet the challenge. As we've seen time and time again, every time we reach a physical barrier, we break through it with technology and innovation. In this case, by replacing copper with fiber, because unlike electrons traveling over copper wires, the distance that photons can carry a signal through glass is largely unrelated to the bandwidth. As AI infrastructure demands even higher transmission speeds and needs to scale to larger and more complex systems, spanning millions of processors woven together now, not thousands or hundreds, optical connectivity will increasingly become the de facto solution. The real question becomes, what does it take to deliver optics across the full AI infrastructure stack? What's it going to take? Well, it starts with recognizing there is no single technology for the entire data center.
It's not how this works. There's no one-size-fits-all solution. There's no shortcuts. There's no easy way to the end here. There's not a single architecture, modulation scheme, frequency band, unique technology that's going to do it all. There's no free lunch. We are pursuing a bunch of different unique optical paths across every distance to get here. Each one of these technologies that I have up here is optimized for a different design point. Each one enables a critical part of the infrastructure in addressing different requirements for density, bandwidth, power, and integration all across the stack. If optical interconnect is the underlying technology for which next-generation AI infrastructure is built, then Marvell is building the broadest portfolio with the deepest bench in the industry. No company can deliver this transformation alone. As Jensen talked about earlier, it takes an ecosystem to get here.
Like I said, technology innovation is great. It's part of the challenge, but not all of it. Demonstrating this at scale is really what matters. At this point, if you're just operating on a PowerPoint or a demo, a POC, a press release, it's not going to get you there. Customers need solutions now that are ready, they're reliable, they need to be manufacturable, and be ready to deploy at scale. Marvell and our ecosystem partners have been doing this for a long time. We've already shipped hundreds of millions of DSPs. We've accumulated through our volumes, tens of billions of device hours of data in the field. This experience matters because these products have to work, not just in the lab, but in the world's largest data centers at very high volume and very reliably for years. That requires investing ahead in the manufacturing ecosystem.
You've got to build the capacity and the supply chain infrastructure before the market arrives. This is why the ecosystem matters so much, and it matters a lot here in Taiwan, by the way. One of our most important partners at Marvell in this journey has been Advanced Semiconductor Engineering or ASE. ASE is one of the world's leading semiconductor manufacturing companies. They have more than 100,000 employees with operation in Asia and actually all around the globe, with a decades-long track record of helping enable pretty much every major technology transition we've gone through in the semiconductor industry. Leading ASE through this period of transformation is someone that I know quite well. He spent more than 25 years helping shape both the company and the industry. Today, I'm thrilled to have my next guest speaker come up, which is ASE CEO, Dr. Tien Wu.
Tien, please join me on the stage. Thank you. Hey. Tien, how are you?
Thank you for inviting me to COMPUTEX.
Great to see you.
Yeah.
It's an honor to have you on stage with us.
Well, it's my honor.
Look, we've been working together a long time.
Yeah.
When I became the CEO, we had a set of ambitions. We talked to a lot of our suppliers. I've known you even before I was the Marvell Technology CEO, when I was an executive back at Maxim, and we worked together there. Part of what, and maybe explain to the audience, too, that sometimes people don't realize, is that as a key supplier into this ecosystem, you have to make bets. You've got to make bets on the companies you work with. You've got to make bets on who you think is going to be successful. We really appreciate that ASE bet on Marvell very early. We've seen great success actually based on that. I'm just curious if you could share your perspective maybe on where Marvell was, what your thought process is, and then where are we today in our journey together.
It'd be great to hear from you, Tien. Thank you.
Okay. I think the best way to describe this is a gradual process. The first decision was not difficult. Marvell, fabless company, has a very good reputation, has gone through a lot of transition. The track record of Marvell has already been there. The product set was a little bit obsolete at the time when you joined. The first one is the business model needs to be aligned. Taiwan ASE is in the manufacturing sector, so we're looking for bet, not only on betting on your success, we're also betting on somebody who can provide the insight for the next generation architecture and also the technology requirement. As you know, the Taiwan company invest infrastructure and CapEx 10 years ahead of time. Big bet. We're only counting on whatever capacity we put it in will be needed and will be utilized.
Right.
That's how we make money.
Right.
Betting on company that we believe will give us very good insight well into the future becomes very important. That's how the decision was made at the very beginning. For the last 10 years, I'm just really happy. Everything that we talk about, it was a dream 10 years ago.
It was a dream.
Today, we are going to ship it. You just mentioned that you're going to have 40% growth for the next few years. I believe you're going to be that.
Right.
We're busy now preparing the capacity for you.
Yes.
We also appreciate that over the last 10 years, we have gone through a lot of strategic discussion. Right? You make commitment to us, we make investment for you. Over time, we're going to produce more of your parts. I think that's the really short story for how that decision was come.
Yeah, no, it's been a great story. Maybe one more for you. The ecosystem here in Taiwan is so unique, and like you said, it takes a decade of investment before you really can see the return, and there's just such a power that's happening here. How do you describe it to people here? Also, there's a lot of people around the world watching. What makes it possible here? Why is it unique? What also makes it difficult to replicate this in the rest of the world?
Okay.
At the same time, there's globalization, so how do we think about those dynamics? I think that'd be an interesting one.
I think the reason why you're asking the question is there's a lot of competing forces and also uncertainty across the world. I think my belief is any business needs to have vision as well as long-term alignment on value. In the business model, the whole Taiwan sector is built on capacity utilization and also innovation, technology investment way ahead of the curve. That's what Taiwan's value. With the fabless company or with specific IDM company, that business model aligns. Beneath that will be the economy of scale. Taiwan accumulated 40 years based on the PC transition to the wireless, to the mobile computing, to the data center. Now we're into HPC. That 40 years of experience accumulated 350,000 semiconductor employee, also accumulated 1.1 million high tech employee, and many of them are here.
That experience becomes extremely valuable combined with the economy of scale as well as the cluster efficiencies. When you think about the workforce with years of experience behind it, when you think about a cluster efficiency, when you think about the capacity, economy of scale, we already put it in. One more thing, I think Taiwan, good or bad, we had fewer choices than the other region, like United States. Most of the engineer, when they come out, they have few choices to make. Semiconductor, IT industry becomes attractive choices in Taiwan, not necessarily in the other region. With all of that combined, I think this ecosystem is very difficult to replicate. It is not impossible, but will take years.
Right. Great. Well, thank you so much.
Okay. Well, thank you.
I appreciate the partnership so much.
No, I appreciate it. Thank you.
We're off to the races, Tien. Thank you. Tien Wu.
Thank you.
Okay, so like we said, the future of AI data centers is all optically connected infrastructure, and you heard him say it. This is going to drive a tidal wave of growth, innovation that's needed, and scale in manufacturing. What does that inevitable future actually look like? If you just take a step back for a minute and you actually don't think about right now, think about 10 years in the future, and it's a world where a lot of the copper connections are gone, and just think about a world where data transmission now, at some point, is all optical. This is a world where then distance doesn't matter, actually, and that's a profound change. Servers, racks, and overall data center architectures today have all been designed around the constraints of distance, and software workloads actually have also been optimized around those same constraints.
What if distance no longer matters? How might the architecture itself change? What new capabilities become possible when the infrastructure is no longer constrained by distance? Let's start with the scale-up network in the rack. As we discussed earlier, this is where we can connect the largest possible number of processors together in a full any-to-any configuration. In the past, the size of this domain was limited by the length of the copper connection, but with optics, distance doesn't matter. Now we can change the size of the scale-up domain from 72 or 144 XPUs or GPUs to 1,000 or more, all optically interconnected. The implications for workloads are enormous. Today, AI workloads must be broken down into smaller sub-problems that fit within the scale-up cluster, because communicating outside the cluster today is slower, much lower bandwidth.
Optically interconnected systems can manage workloads on an order of magnitude larger, and it does not stop there, by the way. What happens when the optical connectivity comes inside the server itself? Modern AI servers are composed of a certain number of CPUs, XPUs, memory, and network interfaces. The reason they're all in the same system is because of distance. The CPUs and XPUs need to access memory at very high bandwidth, which means they need to sit right next to each other on the board with copper traces serving as the connections between them. In a future where these connections are all optical, distance actually doesn't matter. You can imagine a completely disaggregated architecture, XPUs in one system, memory in another, agentic CPUs in another, which unlocks another possibility. In today's systems, the ratio of CPU and XPU or GPU, it's fixed.
These ratios have to be defined at the time the system is built and deployed. No two workloads require exactly the same ratio. Jensen talked about this, actually. Which means at any given time, some portion of the computer memory could be underutilized for a given workload. That costs money. Once we decompose the system into separate pools of compute memory and they're all optically interconnected, we can then decompose dedicated systems on the fly, which are then optimized for whatever the workload is. Imagine future data centers, a globally optically interconnected data infrastructure. These rigid boundaries we have today and the systems we have, they begin to disappear. Compute can now be pooled, memory can be pooled, and infrastructure can be composed dynamically at scale.
For the first time, architects can begin designing AI systems around the needs of the model, not around the limits of the interconnect. This is where AI infrastructure is headed. It's a data center without distance, where compute, memory, networking, and photonics operate as one unified system. Where millions of resources across the data center can work together as if they were one machine. An architecture defined by the needs of the workload, not the limits of the connectivity. We believe this is the next era of computing infrastructure, and Marvell is helping build the connectivity foundation that will make all this possible. Thank you very much for your time today.