Hot Chips 2026: NVIDIA details the 88-core Olympus Vera CPU

In one sentence At the Hot Chips conference at Stanford (August 23-25) NVIDIA gave the first architectural deep dive of its Vera CPU: 88 custom Olympus Arm cores on a monolithic die, spatial multithreading, LPDDR5X memory with 1.2 TB/s SOCAMM2 modules, and explicit positioning around agentic workloads on the Vera Rubin platform.

Needs review Reputable source

ShareLinkedIn X

At Hot Chips, the annual Stanford gathering where chipmakers reveal their most anticipated designs, NVIDIA opened the hood of Vera for the first time, the processor that will sit next to its upcoming Rubin GPUs in AI datacenters. Until now only its promises were known; now the technical details are on the table.

The core design choice tells you where AI is heading. Instead of chasing the highest possible core count, as server tradition dictates, NVIDIA bet on 88 individually fast cores. The reason: modern AI agents alternate GPU work with constant orchestration, tool calls and data preparation, tasks where single-core speed matters more than sheer quantity.

NVIDIA claims 1.8x faster completion of agentic tasks and twice the efficiency of traditional x86 CPUs, and on stage showed head-to-head comparisons with an AMD EPYC on Linux kernel compilation. Vendor numbers, pending independent verification, but the message to datacenter operators is clear: in the Vera Rubin platform even the CPU is designed around agents.