Understand this story
On November 5-6, 2026, NVIDIA announced and released the Vera CPU for AI agent workloads, claiming that in a 100-megawatt data center comparison, a single Vera outperforms the fastest x86 CPU and the most energy-efficient x86 CPU. NVIDIA's figures are: compared with the fastest x86 CPU, Vera can support 1.6 times the number of agents per watt; compared with the most energy-efficient x86 CPU, agent speed is 2 times faster. Early adopters including Perplexity, DeepInfra, ClickHouse, DuckLabs, Daytona, and Modal provided production environment test results, among which Daytona used Vera to launch 1,000 agent sandboxes in 10.8 seconds, versus 88.4 seconds on x86. According to a separate Unite.AI report, Perplexity said that in early tests, Computer-type workflows were about 1.5 times faster than the existing production reference, and concurrent sandbox startup was up to 1.9 times faster, and it plans to deploy Vera in subsequent production systems; DeepInfra measured 29 milliseconds for the bare-metal orchestration step, versus 35 milliseconds for AMD Zen 5. Vera's specifications are 88 Olympus cores and 176 threads per chip, up to 1.2 TB/s LPDDR5X memory bandwidth, and 1.5 TB of memory.
AI-generated from public reports
Latest development · Report
NVIDIA Publishes Vera CPU Agent Performance Data, Up to 2x Faster Than x86Developments
1 confirmed developments
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 6
Unite.AI
NVIDIA Publishes Vera CPU Agent Performance Data, Up to 2x Faster Than x86
On November 5, 2026, NVIDIA published Vera CPU test results, claiming that in a 100-megawatt data center comparison, agent speed is 2x that of the most energy-efficient x86 CPU, and agents per watt is 1.6x that of the fastest x86 CPU. Perplexity said that in early tests, Computer-type workflows were about 1.5x faster than the existing production reference, concurrent sandbox startup was up to 1.9x faster, and it plans to deploy Vera in subsequent production systems. DeepInfra, Daytona, ClickHouse, and DuckDB also provided comparison data; among them, DeepInfra measured 29 milliseconds for a bare-metal orchestration step, versus 35 milliseconds for AMD Zen 5; under secure microVMs, Vera sustained 256 concurrent agents, versus 160 for x86. Vera's specifications are 88 Olympus cores and 176 threads per chip, up to 1.2 TB/s LPDDR5X memory bandwidth, and 1.5 TB of memory.
NVIDIA BlogFirst-party
NVIDIA Releases Vera CPU, Focused on Agent Throughput and Energy Efficiency
NVIDIA has released the Vera CPU for AI agent workloads, claiming that in a 100-megawatt data center comparison, a single Vera outperforms the fastest x86 CPU and the most energy-efficient x86 CPU. NVIDIA's figures are: compared with the fastest x86 CPU, Vera can support 1.6 times the number of agents per watt; compared with the most energy-efficient x86 CPU, agent speed is 2 times faster. Early adopters including Perplexity, DeepInfra, ClickHouse, DuckLabs, Daytona, and Modal provided production environment test results, among which Daytona used Vera to launch 1,000 agent sandboxes in 10.8 seconds, versus 88.4 seconds on x86.