Polylog
The Polylog AI Intelligence Brief

Morning Edition · Wednesday, July 22, 2026Published at 1:47 AM EDT · New York

Nvidia Moves Vera Rubin Into Production as CoreWeave Reports 10x More Tokens Per Megawatt

The NVL72 rack passed a 147-hour validation suite at CoreWeave, which measured a tenfold gain in tokens per megawatt over Grace Blackwell running DeepSeek-R1.

Nvidia Moves Vera Rubin Into Production as CoreWeave Reports 10x More Tokens Per Megawatt

Nvidia said its Vera Rubin NVL72 platform is ramping into production, with racks deploying at CoreWeave, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure and broader shipments planned for the second half of 2026. Each rack pai…

Continue the AI Intelligence Brief

Track frontier labs, chips, export controls, model releases, regulation, and AI infrastructure.

  • 5 AI intelligence signals a day
  • Frontier labs, compute, and chips
  • Model releases and AI infrastructure
  • Source-grounded analysis with confidence labels

The Global Intelligence Brief stays free.

Part of a tracked trend

The Inference-Cost Efficiency Race

Techniques that cut tokens generated and KV-cache memory per query will keep compressing the marginal cost of serving reasoning models, making inference efficiency a recurring competitive axis alongside raw capability.