Semiconductor News & Analysis Feed

__fail__

__fail__

Nscale announced an expansion of its collaboration with Microsoft and Start Campus that will deploy more than 66,000 NVIDIA Rubin GPUs at the SINES Data Campus in Portugal beginning in late 2027.The deployment of NVIDIA Vera Rubin NVL72 systems is expected to become one of the largest AI infrastructure installations in the European Union and will support Microsoft’s European customers with frontie

A new technical paper, “AMMA: A Multi-Chiplet Memory-Centric Architecture for Low-Latency 1M Context Attention Serving,” was published by researchers at UC San Diego, Columbia University, Yonsei University, NVIDIA, and Samsung.Abstract“All current LLM serving systems place the GPU at the center, from production-level attention-FFN disaggregation to NVIDIA’s Rubin GPU-LPU heterogeneous platform. Ev

__fail__

LONDON, UNITED KINGDOM— 5 May 2026 — Nscale is announcing an expansion of its collaboration with Microsoft and Start Campus with 66,000+ NVIDIA Rubin GPUs, starting in late 2027. This agreement builds on the deployment of over 12,600 NVIDIA Blackwell Ultra GPUs at the first building of the SINES Data Campus for Microsoft. As part of the agreement, Nscale’s additional investments of €230 million i

SambaNova used a Computex 2026 session on June 4 to make its most public case yet that the GPU-only approach to AI inference is hitting a fundamental wall — and to demonstrate, live on stage, an alternative architecture it calls disaggregated inference running in a production data center.