ai-hardware

Samsung begins commercial HBM4 shipments with 3 TB/s bandwidth per stack, gaining a timing edge over SK Hynix as memory becomes the critical bottleneck in AI infrastructure.

Reno-based Positron AI has raised $230 million at a $1B+ valuation, claiming its Atlas chip matches Nvidia's H100 inference performance at one-third the power while its next-gen Asimov silicon promises six times the memory of Nvidia's Rubin.

NVIDIA announces its most ambitious AI platform yet at CES 2026, integrating six new chips designed to dramatically reduce the cost of running AI models.