Nvidia is planning a new class of GPU called Rubin CPX, designed specifically for the compute-heavy analysis phase in AI models. The strategy, known as split inference, is backed by new benchmark records from Nvidia’s Blackwell Ultra architecture, which uses a similar approach in software. The article Rubin CPX is Nvidia's first GPU built specifically for massive-context AI applications appeared first on THE DECODER.
Rubin CPX is Nvidia’s first GPU built specifically for massive-context AI applications
Nvidia is planning a new class of GPU called Rubin CPX, designed specifically for the compute-heavy analysis phase in AI models. The strategy, known as split inference, is backed by new benchmark records from Nvidia’s Blackwell Ultra architecture, which uses a similar approach in software. The article Rubin CPX is Nvidia's first GPU built specifically for massive-context AI applications appeared first on THE DECODER.