RAISE Summit '26 | Mark Papermaster, AMD

Clips
More from RAISE Summit '26

Mark Papermaster

CTO & EVP

AMD

play_circle_outline Agentic AI driving heterogeneous, system-level compute architectures

play_circle_outline Agentic AI driving heterogeneous, system-level compute architectures across CPU, GPU, network, memory.

play_circle_outline AMD's System-First Push: Xilinx, Pensando, ZT Systems Reduce Data-Center Latency and Enable Rack-Level Networking

play_circle_outline AMD's System-Optimizer Shift: Low-Latency CPU–GPU Memory and Networking, Rack-Level Advantage via Xilinx, Pensando, ZT Systems

play_circle_outline Helios rack featuring 6th gen EPYC and MI455X for training clusters

play_circle_outline Cost-Effective AI: Inference vs Training Economics, Policy-Based Routing, and ROCm Portability from Cloud to Edge and PCs

play_circle_outline AMD’s Modular Compute Stack: Helios Rack, 6th‑Gen EPYC, MI455X and MI350P from Supercomputers to Edge

play_circle_outline Constraint-Driven Design: MI350P PCIe Scales Big GPUs to Mid-Size Servers with Modular CPU, GPU, Neural Integration

Info

Mark Papermaster, AMD

Mark Papermaster

CTO & EVP AMD

search

John Furrier

>> connecting Silicon Valley and Wall Street. I'm John Furrier, the co-host of theCUBE here with Dave Vellante, my co-host.

>> Welcome to the RAISE Summit coverage, brought to you by Solidigm, with support from Tensordyne, d-Matrix, and Argentum AI.

John Furrier

>> Welcome back to theCUBE's live stream in Paris, France, for the RAISE Summit.

John Furrier

>> I'm John Furrier, host of theCUBE. Day one kicking off. Day zero, sidecar event, yesterday was MACHINA. Physical AI and robotics, which is going to lead into a hot year in robotics, but that's all part of the power and the computing and the software efficiencies coming from all these new integrated systems. This next guest is a legend in the industry, the conscious engineering brain for AMD, 40 year history Mark Papermaster, CTO and EVP with AMD. Thanks for coming on theCUBE. Good to see you again.

Mark Papermaster

>> John, great to see you.

John Furrier

>> CUBE alumni back. I say you're the conscious brain of AMD because the engineering mindset of AMD has been one of the hallmarks of the company. And there's so much you guys are doing. It's so broad in the spectrum of capabilities, but performance has been a very huge deal. So thank you for what you do. We appreciate more horsepower, go faster. What's your current take right now? Because right now the demand is so high for performance. But it's not the way it was last decade, last generation. it's transitioning to a new architecture, new requirements, we're paving new roads, we're pouring new concrete in AI infrastructure, what's your take on all this?

Mark Papermaster

>> Well I love your analogy, we're paving new roads. John, so true, because look, the need for high performance, that's actually not new. So I'm four decades plus in the industry, so that means four decades plus driving more and more performance and efficiency across the compute infrastructure. So that's been my whole working career. But it is a new road now, because agentic AI is changing everything. It's such a diverse workload, John, that it's driving us at AMD, and certainly the whole industry, to think, design, architect, and implement much more holistically than we ever have. And what I mean by that is, the workloads are so complex because people are looking at what they do end to end. Again, they're looking at whole processes, not just one bespoke task. And that means you need different computing engines, and they need to work together. And they need to work together at scale. We're talking across massive clusters of racks.

John Furrier

>> I love that analogy of focusing here, and now we're looking at a larger scale, because at the end of the day, it's all a system. So if you have a system mindset, what's changed the most for you as you evaluate efficiency and some of the constraints and managing through those because when you look at the broader workloads that are running on Instinct, there are use cases, there are new requirements. What are some of those system elements that need to be in place as first principles? That's different than building a motherboard and putting software on it, accessing it. Now we got, and then putting it into a data center, then putting it in the cloud. What's different now from a system standpoint from a first principles?

Mark Papermaster

>> Well, when you look at data center, the movement of the data needs to be so efficient. Because it's always been moving data from memory into that beast, that processor core, moving it from I/O into that processor core as efficiently as you can. But again, what I said a moment ago, it's heterogeneous now. So now you have to think about that data movement and optimizing that performance across CPU, across GPU, and across the network, and all of that to a memory complex, you need low latency to memory, you need to get out to storage, and so what's changed is not the local optimization on each of the elements, you still have to do all of that,

Mark Papermaster

>> but you have to, as you said, at a system level, really optimize, and that's what we've changed at AMD.

Mark Papermaster

>> We've become system optimizers, still optimizing every component, but then equally looking at how you optimize how each of the pieces come together. And in that vein, we've expanded our portfolio. So look at what we've done over the last few years, right? We added networking capability, we acquired Xilinx, we acquired Pensando, and then we acquired ZT Systems that brought a whole rack design capability.

Mark Papermaster

>> So for the data center, John, we're optimizing at the rack level, hardware and software.

John Furrier

>> All right, so I have to ask you this question

Mark Papermaster

>> because it comes up a lot.

John Furrier

>> Okay, I've got this large scale system, I want to put this AI factory or refinery into the system.

John Furrier

>> Great, I plug it in, I put it into a data center, great. But whoa, I got all this other stuff. So backwards compatibility is a big discussion that's come up a lot on theCUBE, but also backwards compatibility with existing workloads. So I have a lot of AMD, but I might want to run at the edge. I'm going to have distributed computing architecture. Now I'm looking at a cost envelope and a performance envelope. So I might look at, I might get some performance here, but if I look at the big picture, the performance cost ratio changes. A lot of these infrastructures have a lot of AMD, a lot of x86.

Mark Papermaster

>> That's right.

Mark Papermaster

>> What's your vision on that? Because this has come up a lot.

Mark Papermaster

>> Yes, it's coming up, and you say a lot, I'll say for me, every enterprise conversation, John, I'm having now, this comes up. Because enterprises have, for decades they've been running x86. They're not going to move off that install base. and what we've done at AMD since we launched the new Zen processor back in 2017, we're now on our sixth generation. So at our Advancing AI event on July 22nd and 23rd, we're rolling out this new generation. It continues the kind of leadership x86 CPU, but it's designed in such a way that it matches what I just described a moment ago. It's optimized for standalone x86, traditional workloads, but what about when you're running these agentic workloads, and you need a GPU, but then you still have your traditional x86 apps?

Mark Papermaster

>> Helios has that sixth generation EPYC processor built in.

Mark Papermaster

>> And so they're at the core of the Helios rack with our new MI455X that, again, we'll be sharing more details in a couple weeks.

John Furrier

>> We'll be there, by the way, we're streaming live from your event.

Mark Papermaster

>> Excellent.

Mark Papermaster

>> So we'll be there.

John Furrier

>> Let's get to the price performance, because what you're getting into is, obviously the efficiency of your architecture does that. Performance is obviously great. But the cost also is important. Price performance has always been a benchmark.

John Furrier

>> Let's map that to the current environment. So x86 factors into cost. What about when you look at things agentic workloads spanning to the edge, and inference needs that are different in scope. So I might say, what's the weather like in Palo Alto?

Mark Papermaster

>> Of course.

John Furrier

>> That's different than saying, hey, solve my high-end math problem and write this new thesis and solve these big problems. They have different levels of differentiated services. So I'd love to have that route in the right areas. How does that impact you guys? Because that's coming up a lot.

Mark Papermaster

>> Basically, cloud-based resources are becoming a big cost factor.

Mark Papermaster

>> Yeah, no, you're spot on.And that's what we've architected in our stack. So our AI stack's open. It's called ROCm. It's the same ROCm that we run on those massive clusters. When you're running AI at the edge, you're running AI, local AI on a PC like you have in front of you there, it's also ROCm. And so if you think about what we're doing, we're making it very easy to port. Now what you have to do is, you have to orchestrate your biggest task, a massive inference model that's running on a foundation model, that's running in the cloud or a big data center. But most enterprises, that's very expensive if you run everything there. So they're exactly on the point you made, John. They're looking to run that more economically, and often at the edge it has to be done locally because you need real-time response. And so we've done that for not only our CPU and GPU, but the embedded neural processors that we have on the PCs, and also in the embedded edge.

John Furrier

>> So more flexibility for...

Mark Papermaster

>> More flexibility.

John Furrier

>> All right, so I want to bring something else up. It's not related to the question I'm going to ask, but it's a way to tease out the question. So if you look at what's going on in China, they get constrained GPUs, and so they're smart engineers too out there, and what they do is they actually figure out how to run better, faster. They cluster it together.

John Furrier

>> All entrepreneurial engineers will figure out how to work around constraints.

John Furrier

>> So they're constrained, they don't have the big GPUs, so what do they do? They engineer better solutions to run on non -GPUs. I think that's going to be something that's going to...

Mark Papermaster

>> Or even just lower capability GPUs.

John Furrier

>> So I think that's a tell-tale sign of how the engineering entrepreneurial thinking's going to happen, the constraints are going to be in memory. We're going to see some constraints around cost and deployment, savvy, IQ. I think this is a topic. What are your thoughts on this engineering around constraints as they will come up? Because we're still at the beginning of a new era. What are some of those thoughts around what will come? Because an edge box might not have as big a footprint as a big monster data center.

Mark Papermaster

>> Right.

John Furrier

>> But they'll talk to each other.

Mark Papermaster

>> Right. But tokens will be running in both.

Mark Papermaster

>> Exactly, but that's where modularity comes into play.We've designed our portfolio. We have the broadest portfolio in the industry because of those acquisitions I described and because of our history across data center, PCs, and embedded. And so the problem you described,

Mark Papermaster

>> that's exactly been our focus. Leverage the broad portfolio and enable it. So constraints are the mother of invention. and what we've done is take our big AMD Instinct MI GPUs.

Mark Papermaster

>> Those run in massive clusters. We power the most powerful supercomputers in the world.

Mark Papermaster

>> But we now scale that down into the new MI350P as a PCIe card you can plug right into an x86 server.

Mark Papermaster

>> So that was necessity that drove that. Our customers said, look, I can't run all my workloads on these massive clusters. I'm a mid-sized enterprise, what do you got for me? And so we do have that solution. And we take that then across to the neural processor, right into the PC, right into the edge, so it is in fact that customer demand driving invention, and what we've done is created a modularity intoour portfolio to enable it.

Mark Papermaster

>> I love having you on theCUBE with all your experience and your leadership and knowledge.

John Furrier

>> So I have to ask you, it's like having a consultant sitting at the table with me, because I'm seeing this, this is a question I'm seeing. You guys are building as much capability to manufacture semiconductors and chips and software as fast as possible. The demand curve is undeniable, people see the demand curve. So you're doing your job, and the speed of building is a goal.

John Furrier

>> Outside, your customers are trying to build large-scale infrastructure, housing, data centers, footprint, as fast as possible. So there's a demand there. And so people are trying to go faster, your customers. And there's this new model emerging, when we just talked to Cerebras, they've got a chip, big chip. I talked to a hot startup, Argentum AI, they're just trying to build. So you have this kind of general contractor mindset. The demand is so big they can stockpile the demand and just get the best supply possible. So there's a lot of builders. And they're building large scale systems fast, and they want the best product. What's your advice to them, because they're also thinking the same thing that you're thinking about, how do I get this up and running as fast as possible? They're almost like the next Michael Dell is out there, but they're building systems.

Mark Papermaster

>> Sure.

John Furrier

>> At large scale, and they got to move fast. What's your advice to them?

Mark Papermaster

>> Well, my advice to all these customers is, really understand the problem you're solving. if you look at, if you need a wafer scale, you might be using that because you need real low latency. You might have vibe coding you're doing where you just want that immediate response.

Mark Papermaster

>> You might be most interested in total cost of ownership, just running most efficiently. And more than likely, these days, you have a mixture of workloads. You need multiple optimized systems. So John, the days of the homogeneous data center and homogeneous compute, where a CIO could just say, I'm going to just populate my compute with one engine, one vendor, that's gone.

John Furrier

>> Yeah, and I saw an example where someone was, I won't say the name, but they were kind of complaining to me or crying to me about, yeah, we spent all this money for training. Yeah, now it's all moving to inference. Well, we did that already. It's not a lot of, well, there's still demand for training, but not at the scale of inference. Right. And inference is where the token economy is going to live.

Mark Papermaster

>> Right, no, you need both, and we support both, training and inference.

Mark Papermaster

>> The MI455X cluster we're coming out with is a massive training cluster. But the demand for inference we saw was taking off. So we optimized configurations that really provide efficient inferencing. And if you need the density, that's the big cluster. But then again, with our modularity, you can run that on premise, you can run it air cooled.

John Furrier

>> Modularity and the x86, I think the x86 is going to be a big factor for what I call inference engines that'll be deployed. Okay, what's the coolest thing you're working on? You've got the big event this month. You're going to have a lot of stuff, but can you share what's exciting you these days? It's a great time to be an engineer.

Mark Papermaster

>> It is. I pinch myself sometimes.Well, what I have to tell you is, the last, I'll call it six, seven months, what we've seen is agentic AI is transforming how, even at AMD, we're running engineering. And it's going to transform every enterprise. So we're looking at every process we had. We're saying, look, we can create agentic workflows. We can tie these together. We're shaving months off of our chip design schedule. We're going to be able to get to market faster. And our story applies to the kind of chip design we do, but it's going to be repeated at every enterprise. It's causing everybody to rethink how they do what they do.

John Furrier

>> Yeah, I think everyone wants you to go faster. We love what you do. We're paving new roads, we're pouring new concrete. I think the AI infrastructure is just beginning. if you look at the 5G build out alone, that was close to a trillion. If you look at all the telecom, trillions and trillions of dollars. The upside for AI infrastructure is incredible.

John Furrier

>> And again, it's a shift, the shift has happened.

Mark Papermaster

>> The shift is happening right now. And again, we saw this coming. we have the highest density of CPUs, the most efficient CPUs, the highest performance CPUs, and they're x86, they run your application. And we complement that with the GPUs, our network infrastructure, and our rack optimizations, through PCs, through embedded. John, we could not be more excited. We feel like we introduced the right products, optimized products, right as this wave is completely taking off.

John Furrier

>> Well, we appreciate you. Thanks for coming on theCUBE. Thanks for spending the time out of your busy schedule. CTO and EVP at AMD, efficiency, price performance, but the new architecture, the new workloads are setting a new architecture for the infrastructure and then the applications that are needed.

John Furrier

>> We're hearing it on stage, Lovable, Replit, all the applications, coding, you name it. And we didn't even get to sovereignty yet, but a lot more of that coming here. I'm John Furrier, your host of theCUBE. Thanks for watching.

Mark Papermaster

>> Thank you.

Mark Papermaster, AMD

search

John Furrier

>> connecting Silicon Valley and Wall Street. I'm John Furrier, the co-host of theCUBE here with Dave Vellante, my co-host.

>> Welcome to the RAISE Summit coverage, brought to you by Solidigm, with support from Tensordyne, d-Matrix, and Argentum AI.

John Furrier

>> Welcome back to theCUBE's live stream in Paris, France, for the RAISE Summit.

John Furrier

>> I'm John Furrier, host of theCUBE. Day one kicking off. Day zero, sidecar event, yesterday was MACHINA. Physical AI and robotics, which is going to lead into a hot year in robotics, but that's all part of the power and the computing and the software efficiencies coming from all these new integrated systems. This next guest is a legend in the industry, the conscious engineering brain for AMD, 40 year history Mark Papermaster, CTO and EVP with AMD. Thanks for coming on theCUBE. Good to see you again.

Mark Papermaster

>> John, great to see you.

John Furrier

>> CUBE alumni back. I say you're the conscious brain of AMD because the engineering mindset of AMD has been one of the hallmarks of the company. And there's so much you guys are doing. It's so broad in the spectrum of capabilities, but performance has been a very huge deal. So thank you for what you do. We appreciate more horsepower, go faster. What's your current take right now? Because right now the demand is so high for performance. But it's not the way it was last decade, last generation. it's transitioning to a new architecture, new requirements, we're paving new roads, we're pouring new concrete in AI infrastructure, what's your take on all this?

Mark Papermaster

>> Well I love your analogy, we're paving new roads. John, so true, because look, the need for high performance, that's actually not new. So I'm four decades plus in the industry, so that means four decades plus driving more and more performance and efficiency across the compute infrastructure. So that's been my whole working career. But it is a new road now, because agentic AI is changing everything. It's such a diverse workload, John, that it's driving us at AMD, and certainly the whole industry, to think, design, architect, and implement much more holistically than we ever have. And what I mean by that is, the workloads are so complex because people are looking at what they do end to end. Again, they're looking at whole processes, not just one bespoke task. And that means you need different computing engines, and they need to work together. And they need to work together at scale. We're talking across massive clusters of racks.

John Furrier

>> I love that analogy of focusing here, and now we're looking at a larger scale, because at the end of the day, it's all a system. So if you have a system mindset, what's changed the most for you as you evaluate efficiency and some of the constraints and managing through those because when you look at the broader workloads that are running on Instinct, there are use cases, there are new requirements. What are some of those system elements that need to be in place as first principles? That's different than building a motherboard and putting software on it, accessing it. Now we got, and then putting it into a data center, then putting it in the cloud. What's different now from a system standpoint from a first principles?

Mark Papermaster

>> Well, when you look at data center, the movement of the data needs to be so efficient. Because it's always been moving data from memory into that beast, that processor core, moving it from I/O into that processor core as efficiently as you can. But again, what I said a moment ago, it's heterogeneous now. So now you have to think about that data movement and optimizing that performance across CPU, across GPU, and across the network, and all of that to a memory complex, you need low latency to memory, you need to get out to storage, and so what's changed is not the local optimization on each of the elements, you still have to do all of that,

Mark Papermaster

>> but you have to, as you said, at a system level, really optimize, and that's what we've changed at AMD.

Mark Papermaster

>> We've become system optimizers, still optimizing every component, but then equally looking at how you optimize how each of the pieces come together. And in that vein, we've expanded our portfolio. So look at what we've done over the last few years, right? We added networking capability, we acquired Xilinx, we acquired Pensando, and then we acquired ZT Systems that brought a whole rack design capability.

Mark Papermaster

>> So for the data center, John, we're optimizing at the rack level, hardware and software.

John Furrier

>> All right, so I have to ask you this question

Mark Papermaster

>> because it comes up a lot.

John Furrier

>> Okay, I've got this large scale system, I want to put this AI factory or refinery into the system.

John Furrier

>> Great, I plug it in, I put it into a data center, great. But whoa, I got all this other stuff. So backwards compatibility is a big discussion that's come up a lot on theCUBE, but also backwards compatibility with existing workloads. So I have a lot of AMD, but I might want to run at the edge. I'm going to have distributed computing architecture. Now I'm looking at a cost envelope and a performance envelope. So I might look at, I might get some performance here, but if I look at the big picture, the performance cost ratio changes. A lot of these infrastructures have a lot of AMD, a lot of x86.

Mark Papermaster

>> That's right.

Mark Papermaster

>> What's your vision on that? Because this has come up a lot.

Mark Papermaster

>> Yes, it's coming up, and you say a lot, I'll say for me, every enterprise conversation, John, I'm having now, this comes up. Because enterprises have, for decades they've been running x86. They're not going to move off that install base. and what we've done at AMD since we launched the new Zen processor back in 2017, we're now on our sixth generation. So at our Advancing AI event on July 22nd and 23rd, we're rolling out this new generation. It continues the kind of leadership x86 CPU, but it's designed in such a way that it matches what I just described a moment ago. It's optimized for standalone x86, traditional workloads, but what about when you're running these agentic workloads, and you need a GPU, but then you still have your traditional x86 apps?

Mark Papermaster

>> Helios has that sixth generation EPYC processor built in.

Mark Papermaster

>> And so they're at the core of the Helios rack with our new MI455X that, again, we'll be sharing more details in a couple weeks.

John Furrier

>> We'll be there, by the way, we're streaming live from your event.

Mark Papermaster

>> Excellent.

Mark Papermaster

>> So we'll be there.

John Furrier

>> Let's get to the price performance, because what you're getting into is, obviously the efficiency of your architecture does that. Performance is obviously great. But the cost also is important. Price performance has always been a benchmark.

John Furrier

>> Let's map that to the current environment. So x86 factors into cost. What about when you look at things agentic workloads spanning to the edge, and inference needs that are different in scope. So I might say, what's the weather like in Palo Alto?

Mark Papermaster

>> Of course.

John Furrier

>> That's different than saying, hey, solve my high-end math problem and write this new thesis and solve these big problems. They have different levels of differentiated services. So I'd love to have that route in the right areas. How does that impact you guys? Because that's coming up a lot.

Mark Papermaster

>> Basically, cloud-based resources are becoming a big cost factor.

Mark Papermaster

>> Yeah, no, you're spot on.And that's what we've architected in our stack. So our AI stack's open. It's called ROCm. It's the same ROCm that we run on those massive clusters. When you're running AI at the edge, you're running AI, local AI on a PC like you have in front of you there, it's also ROCm. And so if you think about what we're doing, we're making it very easy to port. Now what you have to do is, you have to orchestrate your biggest task, a massive inference model that's running on a foundation model, that's running in the cloud or a big data center. But most enterprises, that's very expensive if you run everything there. So they're exactly on the point you made, John. They're looking to run that more economically, and often at the edge it has to be done locally because you need real-time response. And so we've done that for not only our CPU and GPU, but the embedded neural processors that we have on the PCs, and also in the embedded edge.

John Furrier

>> So more flexibility for...

Mark Papermaster

>> More flexibility.

John Furrier

>> All right, so I want to bring something else up. It's not related to the question I'm going to ask, but it's a way to tease out the question. So if you look at what's going on in China, they get constrained GPUs, and so they're smart engineers too out there, and what they do is they actually figure out how to run better, faster. They cluster it together.

John Furrier

>> All entrepreneurial engineers will figure out how to work around constraints.

John Furrier

>> So they're constrained, they don't have the big GPUs, so what do they do? They engineer better solutions to run on non -GPUs. I think that's going to be something that's going to...

Mark Papermaster

>> Or even just lower capability GPUs.

John Furrier

>> So I think that's a tell-tale sign of how the engineering entrepreneurial thinking's going to happen, the constraints are going to be in memory. We're going to see some constraints around cost and deployment, savvy, IQ. I think this is a topic. What are your thoughts on this engineering around constraints as they will come up? Because we're still at the beginning of a new era. What are some of those thoughts around what will come? Because an edge box might not have as big a footprint as a big monster data center.

Mark Papermaster

>> Right.

John Furrier

>> But they'll talk to each other.

Mark Papermaster

>> Right. But tokens will be running in both.

Mark Papermaster

>> Exactly, but that's where modularity comes into play.We've designed our portfolio. We have the broadest portfolio in the industry because of those acquisitions I described and because of our history across data center, PCs, and embedded. And so the problem you described,

Mark Papermaster

>> that's exactly been our focus. Leverage the broad portfolio and enable it. So constraints are the mother of invention. and what we've done is take our big AMD Instinct MI GPUs.

Mark Papermaster

>> Those run in massive clusters. We power the most powerful supercomputers in the world.

Mark Papermaster

>> But we now scale that down into the new MI350P as a PCIe card you can plug right into an x86 server.

Mark Papermaster

>> So that was necessity that drove that. Our customers said, look, I can't run all my workloads on these massive clusters. I'm a mid-sized enterprise, what do you got for me? And so we do have that solution. And we take that then across to the neural processor, right into the PC, right into the edge, so it is in fact that customer demand driving invention, and what we've done is created a modularity intoour portfolio to enable it.

Mark Papermaster

>> I love having you on theCUBE with all your experience and your leadership and knowledge.

John Furrier

>> So I have to ask you, it's like having a consultant sitting at the table with me, because I'm seeing this, this is a question I'm seeing. You guys are building as much capability to manufacture semiconductors and chips and software as fast as possible. The demand curve is undeniable, people see the demand curve. So you're doing your job, and the speed of building is a goal.

John Furrier

>> Outside, your customers are trying to build large-scale infrastructure, housing, data centers, footprint, as fast as possible. So there's a demand there. And so people are trying to go faster, your customers. And there's this new model emerging, when we just talked to Cerebras, they've got a chip, big chip. I talked to a hot startup, Argentum AI, they're just trying to build. So you have this kind of general contractor mindset. The demand is so big they can stockpile the demand and just get the best supply possible. So there's a lot of builders. And they're building large scale systems fast, and they want the best product. What's your advice to them, because they're also thinking the same thing that you're thinking about, how do I get this up and running as fast as possible? They're almost like the next Michael Dell is out there, but they're building systems.

Mark Papermaster

>> Sure.

John Furrier

>> At large scale, and they got to move fast. What's your advice to them?

Mark Papermaster

>> Well, my advice to all these customers is, really understand the problem you're solving. if you look at, if you need a wafer scale, you might be using that because you need real low latency. You might have vibe coding you're doing where you just want that immediate response.

Mark Papermaster

>> You might be most interested in total cost of ownership, just running most efficiently. And more than likely, these days, you have a mixture of workloads. You need multiple optimized systems. So John, the days of the homogeneous data center and homogeneous compute, where a CIO could just say, I'm going to just populate my compute with one engine, one vendor, that's gone.

John Furrier

>> Yeah, and I saw an example where someone was, I won't say the name, but they were kind of complaining to me or crying to me about, yeah, we spent all this money for training. Yeah, now it's all moving to inference. Well, we did that already. It's not a lot of, well, there's still demand for training, but not at the scale of inference. Right. And inference is where the token economy is going to live.

Mark Papermaster

>> Right, no, you need both, and we support both, training and inference.

Mark Papermaster

>> The MI455X cluster we're coming out with is a massive training cluster. But the demand for inference we saw was taking off. So we optimized configurations that really provide efficient inferencing. And if you need the density, that's the big cluster. But then again, with our modularity, you can run that on premise, you can run it air cooled.

John Furrier

>> Modularity and the x86, I think the x86 is going to be a big factor for what I call inference engines that'll be deployed. Okay, what's the coolest thing you're working on? You've got the big event this month. You're going to have a lot of stuff, but can you share what's exciting you these days? It's a great time to be an engineer.

Mark Papermaster

>> It is. I pinch myself sometimes.Well, what I have to tell you is, the last, I'll call it six, seven months, what we've seen is agentic AI is transforming how, even at AMD, we're running engineering. And it's going to transform every enterprise. So we're looking at every process we had. We're saying, look, we can create agentic workflows. We can tie these together. We're shaving months off of our chip design schedule. We're going to be able to get to market faster. And our story applies to the kind of chip design we do, but it's going to be repeated at every enterprise. It's causing everybody to rethink how they do what they do.

John Furrier

>> Yeah, I think everyone wants you to go faster. We love what you do. We're paving new roads, we're pouring new concrete. I think the AI infrastructure is just beginning. if you look at the 5G build out alone, that was close to a trillion. If you look at all the telecom, trillions and trillions of dollars. The upside for AI infrastructure is incredible.

John Furrier

>> And again, it's a shift, the shift has happened.

Mark Papermaster

>> The shift is happening right now. And again, we saw this coming. we have the highest density of CPUs, the most efficient CPUs, the highest performance CPUs, and they're x86, they run your application. And we complement that with the GPUs, our network infrastructure, and our rack optimizations, through PCs, through embedded. John, we could not be more excited. We feel like we introduced the right products, optimized products, right as this wave is completely taking off.

John Furrier

>> Well, we appreciate you. Thanks for coming on theCUBE. Thanks for spending the time out of your busy schedule. CTO and EVP at AMD, efficiency, price performance, but the new architecture, the new workloads are setting a new architecture for the infrastructure and then the applications that are needed.

John Furrier

>> We're hearing it on stage, Lovable, Replit, all the applications, coding, you name it. And we didn't even get to sovereignty yet, but a lot more of that coming here. I'm John Furrier, your host of theCUBE. Thanks for watching.

Mark Papermaster

>> Thank you.