#ai-inference-performance

[ follow ]
#token-generation-speed #groq-acquisition #gpu-architecture #sram-based-systems #rubin-gpu #massive-context-ai
Tech industry
fromTheregister
2 days ago

Nvidia GTC 2026: What to expect at AI Burning Man

Nvidia acquired Groq's token-generation technology to address performance gaps in AI inference workloads, combining GPU architecture with SRAM-based dataflow systems for improved speed and efficiency.
fromIT Pro
6 months ago

Nvidia hails 'another leap in the frontier of AI computing' with Rubin GPU launch

The new GPU will work "hand-in-hand" with the Vera CPU and be housed inside the latest Nvidia Vera Rubin NVL 144 CPX platform, the company said. This is an integrated NVIDIA MGX system that packs 8 exaflops of AI compute to provide 7.5x more AI performance than NVIDIA's GB300 NVL72 systems. It also promises 100TB of fast memory and 1.7 petabytes per second of memory bandwidth in a single rack.
Artificial intelligence
[ Load more ]