Nvidia Vera Rubin NVL72 delivers 30x more work per watt than GB300 on agentic workloads
nordicinst · x · 2026-08-24
Nvidia released performance data for Vera Rubin NVL72 systems, showing up to 30x higher throughput per megawatt and 35x lower token costs compared to GB300 NVL72 on agentic workloads. Measured using SemiAnalysis AgentX real-world coding trajectories, the results highlight significant efficiency gains as AI agents demand exponentially more tokens for long-context tasks.
Related event: NVIDIA's Vera Rubin Delivers 30x Gains Over GB300 in Agent Workloads(2 posts)→
More from Infra
- Agent Demos Are Traps: Production Success Depends on Infrastructure — EmetInteractive · 2026-08-25
- NVIDIA Confirms Groq LPX Production; Nebius First to Deploy in Cloud — IanAndrewsDC · 2026-08-25
- Nvidia's Financial Moves Raise Questions; Vera Rubin Can't Mask AI Economy Concerns — TiernanRayTech · 2026-08-25
- LiquidAI Partners for Mobile Small Model Benchmarking — JosephJacks_ · 2026-08-25
- Open-weight models claim 62% of Vercel AI Gateway traffic — rohanpaul_ai · 2026-08-25
- Mesh LLM: Distributed AI for pooling local compute — alex_verem · 2026-08-25