NVIDIA 发布用于企业与 PC 端智能体的新开放模型
Nvidia is leaning hard into the trend of enterprise companies wanting to run their own open models in their own data centers and also trying to make on-device, private AI agents a thing.
英伟达正大力押注企业希望在自己的数据中心运行自有开放模型的趋势，同时试图让设备端私有 AI 智能体成为现实。
原文来源：TechCrunch（techcrunch.com）。本页摘要与要点整理自公开来源，版权归原作者所有，完整内容请阅读原文。
Inside Prometheus: Meta’s Journey to a Gigawatt-Scale SuperclusterMeta 揭秘吉瓦级超级计算集群 Prometheus
Meta Engineering · 2026-10-19
Hackers hit AI coding tool OpenCode, running up charges on users’ AI accounts开源 AI 编程工具 OpenCode 平台遭攻击，部分用户 API 密钥被盗
TechCrunch · 2026-10-20
Cutting our inference bill by 68% by being obsessive about optimizationCloudflare 通过提示缓存与模型路由将 AI 代码审查推理成本降低 68%
Cloudflare · 2026-10-20
Hugging Face ❤️ TorchTPU: TPUs, NativelyHugging Face 原生支持 TorchTPU
Hugging Face Blog · 2026-10-20
OpenAI acquires Orchestra BioOpenAI 收购 Orchestra Bio，拓展生命科学研发
OpenAI · 2026-10-20
谷歌开源 Harness 层自我改进框架 Harness²
量子位 · 2026-10-19