Token economics reshape AMD’s enterprise AI

Enterprise AI is entering a new phase as organizations shift their focus from experimentation to production deployments that deliver measurable business outcomes. That transition is bringing AI token economics to the forefront, reshaping infrastructure priorities around inference costs and the flexibility to run AI across hybrid environments.

Hybrid AI is evolving beyond simply deciding whether workloads run on-premises, at the edge or in the cloud. Increasingly, enterprises are also adopting a hybrid approach to AI models, combining frontier and open-weight models based on the needs of each workload, according to Suresh Andani (pictured), corporate vice president of compute and enterprise AI at Advanced Micro Devices Inc.

“A lot of enterprises are now figuring out that not all tasks need frontier models. Frontier models are great, by the way,” Andani said. “If you need the deepest context, if you really need the high concurrency, frontier models have really done well for the enterprises. Not every enterprise AI task needs to run on a frontier model.”

Andani spoke with theCUBE’s John Furrier and Dave Vellante at AMD Advancing AI event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the shift toward distributed AI infrastructure and the growing importance of AI token economics as enterprises scale AI into production. (* Disclosure below.)

Optimizing AI token economics

Not every enterprise AI workload requires a frontier model. Many tasks are better suited to open-weight models running on-premises, where organizations can reduce costs while maintaining greater security, data sovereignty and control, according to Andani.

“As a provider of AI compute, we want to make sure that we are enabling those enterprises to not only run their AI tasks through the frontier APIs in the cloud, but we are providing them infrastructure that they can host open-weight models very efficiently on-prem,” Andani said. “Every enterprise conversation I’m in, that’s a key topic of discussion.”

Enterprises are paying for outcomes, not chips or servers. The value of AI infrastructure is ultimately measured by the business results it delivers — a shift that Andani said is changing how organizations justify AI spending.

“If I’m a big oil and gas company and I have an issue with contract leakage, my outcome that I’m driving is, can I use AI to reduce the dollars I am leaking on contract? That’s what they’re trying to drive,” Andani said. “It’s also true that the TCO is also driven significantly by what’s at the foundational layer. Speeds and feeds matter to bring the end customer value in terms of outcome.”

Some enterprises want to build their own AI agent stacks using platforms provided by OEM and cloud partners. Others prefer AI platforms that provide more of the stack out of the box, according to Andani.

“There are certain enterprises that need pre-built agents, like right out of the box, I need a chatbot agent,” Andani said. “For us really it’s to deliver that whole stack from not just the GPU, CPU and networking chips, the platforms, the control plane for AI and then the actual ISV application layers.”

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

(* 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