NVIDIA reports Vera gains in agent speed and power efficiency

NVIDIA says its Vera CPU can improve agent performance and capacity within power-constrained data centers, reducing the need to combine different processors for speed and density. AI agents alternate between GPU reasoning and CPU execution when searching, querying databases, writing code and taking action. As deployments grow, infrastructure providers must support more simultaneous agents without increasing delays or exceeding available power.

In a 100-megawatt data center comparison, NVIDIA reported that Vera supported 1.6x more agents per watt than the fastest x86 CPU while delivering leading agent speed. Against the most energy-efficient x86 CPU, NVIDIA reported 2x faster agent speed with leading energy efficiency. These results position Vera as an option for operators balancing responsiveness with overall workload capacity.

Early adopters also reported gains in production-style testing. Perplexity ran agentic coding tasks 1.5x faster than on traditional CPUs and plans to deploy Vera in an upcoming production system. DeepInfra reported up to 2.2x speed compared with the best x86 processor for agentic orchestration. ClickHouse said its standard binary scaled near-linearly across 176 cores without Vera-specific tuning and topped the ClickBench hardware leaderboard. Modal reported up to 2x sandbox throughput per core versus comparable x86 hardware, with gains increasing under heavy load.