It looks like Nvidia’s AI inference GPU won’t see the light of day this year, which could actually be good news for PC gamers
Nvidia’s Rubin CPX, an AI inference-focused GPU announced last September, appears to be delayed and will not launch in 2026, according to sources cited by Korean publication The Elec. The CPX was notable as Nvidia’s first processor designed specifically for AI inference, distinguishing it from the company’s broader lineup of AI accelerators. While Nvidia’s Vera CPU and other Rubin GPU variants continue to generate strong orders, the CPX has experienced a significant slowdown with no visible product momentum. The delay may provide relief for PC gamers, as it suggests fewer resources are being diverted toward the specialized AI inference market and potentially keeps Nvidia’s consumer GPU pipeline more focused on gaming performance and consumer applications rather than enterprise AI workloads.