Nvidia unveils new GPU designed for long-context inference
-
最新报道：Nvidia unveiled the Rubin CPX GPU at the AI Infrastructure Summit, engineered for context windows exceeding 1 million tokens. Optimized for large-sequence processing and part of the Rubin series, it supports disaggregated inference infrastructure, enhancing performance in long-context applications like video generation and software development. The chip, set for late 2026 release, follows Nvidia’s profitable streak with $41.1 billion in recent data center sales.
下一篇
BMNR's Ethereum Holdings: A Catalyst for Institutional Crypto Reallocation and Ethereum's Core Asset Status
- Following the 2025 Ethereum ETF approvals, BitMine Immersion Technologies (BMNR) has accum...
24小时热点
热点专题
407232