Nvidia has designed a new class of GPU for massive-context inference, the Rubin CPX, due in late 2026. Purpose-built to speed the million-token applications used to generate video and create software, the Rubin CPX functions as a specialty accelerator, working in concert with Nvidia Vera CPUs and Rubin GPUs packaged inside the upcoming Vera Rubin NVL144 CPX rack platform. “The Vera Rubin platform will mark another leap in the frontier of AI computing,” revolutionizing massive-context AI just as RTX did graphics and physical AI, said Nvidia CEO Jensen Huang. Continue reading Nvidia Says Rubin CPX Inference Accelerator Coming in 2026

Facebook, along with Intel, Broadcom and others, revealed the first steps toward developing an open switch that will rival Cisco’s network hardware. The social networking site’s Open Compute Project (OCP) was announced six months ago, and now it’s making progress toward its goal by receiving switch specifications from its consortium members. The OCP plans to release a specification and reference box for an “open, OS-agnostic top-of-rack switch.” Continue reading Facebook’s New Open Compute Project to Compete with Cisco