Agentic AI compute has become the defining workload of the cloud’s next era, reshaping how hyperscalers, neo-clouds and enterprises architect their infrastructure. As autonomous agents move into production, demand for both GPUs and high-core-count CPUs is surging — and the economics of cloud computing are being rewritten in real time.

The shift echoes the cloud’s own explosive rise, but at a far faster pace. Advanced Micro Devices Inc. is positioning itself at the center of it, positioning itself around an open, full-stack approach to the AI era will convert its cloud momentum into a durable advantage, according to Steve Berg (pictured), corporate vice president and general manager server CPU Cloud Business Group at AMD.

“It’s really been phenomenal to see what’s happened over even just the last — it’s not even 18 months that the agentic AI world has existed,” Berg said. “I’ve been doing this in the cloud for about 15 years, and watching the explosion that’s happening is amazing. The GPU is becoming much more relevant again — actually probably the most relevant element of the infrastructure in the agentic AI era today.”

Berg spoke with theCUBE’s John Furrier and Dave Vellante at AMD Advancing AI 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how agentic AI compute is driving demand across cloud and hybrid infrastructure, AMD’s cloud market gains and the road to its next-generation EPYC “Venice” processors. (* Disclosure below.)

Agentic AI compute drives a heterogeneous hardware era

AMD’s rise in the cloud has been years in the making, built on a steady cadence of server chips that captured design wins from major providers. The company now targets nearly 50% of the cloud GPU market on a revenue basis by year’s end, a position it credits to total cost of ownership rather than raw specifications, Berg noted.

“The products have led to being able to win in the cloud world,” Berg said. “It’s a TCO play. It’s always a TCO play when it comes to cloud. And ensuring that they have the right performance per dollar for their workloads is how we’ve been able to become dominant in the cloud.”

That openness has helped AMD hold ground even as hyperscalers design their own silicon. The broader data center GPU market is forecast to expand at a 15.8% compound annual rate through 2030, and much of that growth is flowing toward suppliers that let customers avoid lock-in, Berg said.

“It’s the enterprises that run in the cloud that want to have that open ecosystem, and they don’t want the lock-in,” Berg said. “They want to be able to run the same software and ecosystem that they do on-prem in a hybrid environment as they do in the cloud.”

The agentic era is fueling demand for a range of chip types — storage, high-performance computing and inference workloads alike — rather than one dominant GPU, Berg noted. AMD’s forthcoming Venice EPYC processor, with up to 256 cores and a claimed 1.7 times performance gain, is built for that task-driven load, even as a tight memory market remains the biggest constraint on how fast the company can ship.

“Venice will be the best EPYC, will be the best agentic processor on the planet,” Berg said. “And I think everyone’s going to put it into their infrastructure to run agentic workloads.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of AMD Advancing AI 2026:

(* Disclosure: TheCUBE is a paid media partner for the AMD Advancing AI event. Neither AMD, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE