In a July 2026 presentation highlighting the ripple effects of the artificial intelligence boom, AMD CEO Lisa Su pointed to a surge in the data center CPU total addressable market (TAM) from roughly $26 billion in 2025 to nearly $220 billion by 2030. This compound annual growth rate (CAGR) exceeding 50 percent points to a shift in how robustly cloud service providers and enterprise data centers are building their IT infrastructure.
Around the same time, IDC released its projection of a $497 billion AI infrastructure market and Gartner reported $822 billion in Data Center Systems spending for 2026. From Su’s perspective, most of the growth in the coming years will be driven by meeting increasing AI-related demands. Going forward, she expects to see an explosive demand for agentic AI workloads, designed for autonomous, multi-step AI decision-making.
"We are seeing a major industry shift from the classic call and response AI that was your traditional ChatGPT to now agent calls, and agents can call more and more agents to do more autonomous tasks for us," said Brayden Mahdavi, go-to-market lead and business manager for NeoCloud within AMD's Compute and Enterprise AI organization, in a video interview with The Forecast, recorded during the July 2026 AMD Advancing AI event in San Francisco.
Mahdavi works at the intersection of cloud providers, infrastructure vendors and enterprise customers. One common theme cuts across all of them:
“Customers want options,” he said.
Customers are evaluating their complete tech stack from the silicon that powers the workloads to the private and public clouds that host them and the virtualization software that ties them together so they can be tightly managed.
A New Class of Cloud Steps In
The tier-one hyperscalers, Amazon, Google and Microsoft, remain critical to AMD's business and are not going away, but Mahdavi explained that current capacity and supply challenges is driving customers to seek new or different options. Like a magnet, many are drawn to neoclouds, specialized AI cloud providers that built their reputations in the GPU space and now find fresh relevance as workloads evolve, according to Mahdavi.
"They're so nimble and lean that they can help productize our technology in a way that is favorable for our end customers who are facing challenges around total cost of ownership and other performance challenges," Mahdavi said.
The rise of these providers comes as enterprises confront a stubborn economic reality. A Gartner analysis predicts that more than 40% of agentic AI projects will fail by 2027 because legacy systems cannot support the demands of modern AI execution. Nutanix has responded with a $250 million strategic partnership with AMD to deliver an open AI infrastructure platform for enterprise inference and agentic workloads.
The infrastructure needed to build and train AI models isn’t optimal for AI inference, explained Mahdavi.
The CPU-to-GPU Ratio Comes Back to Earth
For years, the AI hardware story was almost entirely about GPUs. Enterprises bought them in bulk, sometimes as many as eight GPUs for every one CPU, to feed hungry training runs in frontier lab environments. Mahdavi sees that balancing out.
"What we are going to see very likely over the next few years is a shift back to parity, if you will, of the CPU to GPU ratio," he said.
"Where our enterprise customers were typically buying as many as eight GPUs to one CPU, we're seeing that ratio come down much more to a one-to-one ratio because our CPUs are able to handle much of the orchestration and much of the agent calls, the tool calls that are required for these types of workloads."
The reason is structural, he explained. Agents do not just generate text. They coordinate, orchestrate and call tools, workloads that lean on the CPU rather than the GPU. As agentic AI scales, the CPU reclaims a central role in the stack.
From Training to Inference
The rebalancing tracks a deeper transition in the AI lifecycle. Since the 2023 moment when ChatGPT made AI tangible for the mainstream, the industry's compute spend has been dominated by training cycles, forward passes and back propagation at scale to build ever larger models. The next few years will tilt heavily toward inference, the work of putting those trained models to use.
"Now we want to see the outputs of these models," Mahdavi said. "And as we get to that shift, the demand for compute just continues to soar through the roof."
That surge is pushing enterprises to rethink procurement from the ground up. Rather than buying infrastructure and hunting for workloads to fill it, Mahdavi said, customers are starting with the workload and working backward to the right stack. The result is a sharper focus on optimization, whether through formal FinOps practices or raw business pressure to lower the total cost of ownership.
An Ecosystem Built for Choice
The throughline across every layer is optionality,” said Mahdavi. Customers want the ability to choose the silicon, cloud provider, and virtualization software that best fit their business needs. He pointed out an urgent concern about licensing structures, which have shifted, forcing enterprises to get creative as they rebuild or scale their IT stacks.
Mahdavi said the road ahead is a collective effort.
"Between AMD and our partnerships with Nutanix and with our AI cloud providers, we are very confident that we are going to be able to help our end customers accomplish their business requirements," he said.
For enterprises, it’s about having the right balance or hardware and software capabilities to meet their evolving needs. AI agents are coming, and they require the right resources and manageability.
Related:
- AI Shifts from LLMs to Inference
- AI Trends in 2026: Finding the Right Compute Platform for Each Workload
- AI's Defining Decade: Nutanix Leaders on Inference, Agents, and the Road Ahead
- From Cloud Native to AI-Native: The Evolution of Enterprise Infrastructure
Video transcript:
Brayden Mahdavi: I really sit at the intersection of three different groups. I sit between cloud providers, so this is your kind of typical cloud service provider, then your infrastructure providers, and then end customers within our enterprise. My role really gives me a lot of energy because I hear a common theme across all of these groups, and that is customers want options. I'm really seeing it across three main layers. So the first is really the infrastructure layer, right? This is your silicon choice and optionality across different silicon providers. And then you have within the cloud service provider layer, right? So my traditional tier one hyperscalers, but now we're seeing this new emergent AI specialized cloud provider often referred to as neo-clouds. And the third layer is really your virtualization layer. As many enterprises know, licensing structures have changed as of recent, and that is creating a lot of challenges and a lot of ingenuity around how we reinvent our stack to better serve our business needs.If many enterprises can put their whole fleet on this tier one cloud, albeit Amazon, Google, Microsoft, and that business is critical for AMD. It's not going away anytime soon. Given some of the challenges that we're facing in 2026, capacity and supply are huge challenges. We are really hungry at a market level for a new offering, right? And this is where you are seeing these emergent AI cloud providers that are specialized historically maybe in the GPU space, but many of our partners are seeing with the shift to agentic AI, a bigger need for a CPU. They're so nimble and lean that they can help productize our technology in a way that is favorable for our end customers who are facing challenges around total cost of ownership and other performance challenges. We are seeing a major industry shift from the classic call and response AI that was your traditional ChatGPT to now agent calls.
And agents can call more and more agents to do more autonomous tasks for us. They bring us a lot of value. What we don't see is that they demand a lot of compute resources. What we are going to see very likely over the next few years is a shift back to parody, if you will, of the CPU to GPU ratio. So where our enterprise customers were typically buying as many as eight GPUs to one CPU, we're seeing that ratio come down much more to a one-to-one ratio because our CPUs are able to handle much of the orchestration and much of the agent calls, the tool calls that are required for these types of workloads. The last few years from that ChatGPT moment in 2023, when a lot of people realized this is a real invention that is upon us today, the shift has really been many training cycles within the frontier lab environments.
We're doing a lot of forward passes and back propagation at scale to train up these models, whereas in the next couple of years, it will be very heavily favored towards inference. Now we want to see the outputs of these models. And as we get to that shift, the demand for compute just continues to soar through the roof. Our customers are starting with their workload need and working back to solve what the right infrastructure stack is. We are seeing more optimization in that process, whether it's through a FinOps practice or whether it's through just a straight up business need. It's at a time where we are moving into optimization and how do I lower my total cost of ownership? And between AMD and our partnerships with Nutanix and with our AI cloud providers, we are very confident that we are going to be able to help our end customers accomplish their business requirements.
Ken Kaplan is Editor in Chief for The Forecast by Nutanix. Find him on X @kenekaplan and LinkedIn.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.
Related Articles
