Hardware

Vera Rubin’s first MLPerf preview makes inference the headline
The first Vera Rubin NVL72 preview results in MLPerf Inference v6.1 point to large throughput gains, while also showing why full-stack systems matter.
Chips, data centers, devices, networking, and the physical layer of computing.
03 dispatches
The first Vera Rubin NVL72 preview results in MLPerf Inference v6.1 point to large throughput gains, while also showing why full-stack systems matter.

A new Virginia accountability framework proposes local approval, clearer cost allocation, and stricter energy, water, and backup-power standards for large data centers.

Power, transformers, grid connections, and liquid cooling are becoming the strategic constraints behind the AI buildout.