NVIDIA and OpenAI Launch Open-Weight AI Models for Enhanced Reasoning and Deployment Efficiency
-
最新报道：NVIDIA and OpenAI jointly released open-weight AI models gpt-oss-120b and gpt-oss-20b, optimized for NVIDIA GPUs and CUDA platform. The models deliver 1.5M tokens/second inference speed on GB200 NVL72 using NVFP4 4-bit precision. Available as NIM microservices compatible with FlashInfer/Hugging Face, they enable real-time deployment of trillion-parameter LLMs. This partnership boosts open-source AI adoption through 450M CUDA downloads, supporting applications in healthcare, manufacturing, and generative AI.
下一篇
Crypto Founders Urged to Prioritize Listening Over Fundraising Pitches
- At Blockchain Life Forum 2025, Karnika Yashwant advised crypto founders to focus on listen...
24小时热点
热点专题
418154