Two years ago, AI infrastructure planning centered on training. Then inference took the lead. Now agentic AI—which turns a single request into a shifting chain of retrieval, tool calls, code execution, and results—creates workflows that change how much compute is needed and where, with every execution.

Infrastructure built around a single point in time risks misalignment with evolving requirements. Flexibility is no longer a feature; it is a requirement at every layer of the stack, including the CPUs supporting agentic pipelines.

Enterprises already understand this principle. Databases, virtualization, analytics, web services, technical computing, and AI have long required different system profiles. Agentic AI brings more of those profiles into one workflow.

AMD EPYC 9006 “Venice” Evaluation

In response, AMD has released a white paper evaluating 6th Gen AMD EPYC 9006 “Venice” server CPUs across general-purpose, enterprise, cloud-native, AI, and high-performance computing workloads rather than relying on selected benchmarks.

In SPECrate 2026 Integer testing, the AMD EPYC 9996 server CPU delivers 1.2 times the per-core performance of an Nvidia Vera-based platform and 2.24 times its platform-level performance.

Across enterprise and cloud-native testing—including server-side Java, OpenSSL, MongoDB, Redis, NGINX, and transaction processing—AMD reports gains of 2.4x to 3.7x. Against the Intel Xeon 6980P processor, the AMD EPYC 9996 delivers 1.8x to 3.13x performance advantages across molecular dynamics, materials modeling, and weather forecasting. In a modeled 100-kilowatt rack, it delivers an estimated 3.4 times the throughput of a Vera-based platform.

Strong loaded per-core performance can accelerate latency-sensitive work, while core density supports concurrency and improves rack-level throughput. A processor portfolio lets customers optimize for both without imposing the same compromise across the entire fleet.

“Venice” is one part of a portfolio designed around that principle: four families running on a common software foundation, spanning 8-core edge deployments, 256-core flagship processors, and rack-scale AI host nodes. Each role in an agentic pipeline can use the processor profile that fits it without creating a separate operating environment.

The EPYC 9996 is in production today. Major OEM platforms are on track to launch, and leading cloud providers begin deploying later this year.

Agentic AI will continue adding diversity to workload requirements. The answer to more varied demands has never been less choice.