https://index.zshipu.com/ai001/tags/AI%E8%8A%AF%E7%89%87/ Recent content in AI芯片 on 知识铺的博客 Hugo -- gohugo.io zh-CN Mon, 31 Aug 2026 00:00:00 +0000 https://index.zshipu.com/ai001/post/20260831/Nvidia-35%E4%BA%BF%E7%BE%8E%E5%85%83%E6%89%A9%E5%A4%A7%E8%81%94%E5%8F%91%E7%A7%91%E5%90%88%E4%BD%9C-%E9%94%81%E5%AE%9AAI%E5%AE%9A%E5%88%B6%E8%8A%AF%E7%89%87%E5%85%BC%E5%AE%B9/ Mon, 31 Aug 2026 00:00:00 +0000 https://index.zshipu.com/ai001/post/20260831/Nvidia-35%E4%BA%BF%E7%BE%8E%E5%85%83%E6%89%A9%E5%A4%A7%E8%81%94%E5%8F%91%E7%A7%91%E5%90%88%E4%BD%9C-%E9%94%81%E5%AE%9AAI%E5%AE%9A%E5%88%B6%E8%8A%AF%E7%89%87%E5%85%BC%E5%AE%B9/ Nvidia 35亿美元扩大与联发科合作，锁定AI定制芯片兼容 Nvidia以35亿美元扩大与联发科的合作，直接指向AI生态兼容定制芯片。一位投资经理称，此举显示Nvidia正努力让AI系统适配更多自定义硬件，而非只推自家GPU。这笔交易可能帮助其在AI领域获得更广市场，同时也凸显定制芯片趋势 https://index.zshipu.com/ai001/post/20251008/amd-rubin-redesign-nvidia-battle/ Wed, 08 Oct 2025 04:19:31 +0000 https://index.zshipu.com/ai001/post/20251008/amd-rubin-redesign-nvidia-battle/ 据SemiAnalysis最新报告披露,AMD已针对英伟达的下一代Rubin AI芯片进行重大重新设计,在双方为争夺AI加速器市场主导权而展开的"军备竞赛"中再度升级。英伟达将Rubin VR200 GPU的内存带宽从13TB/s大幅提升至20.5TB/s,同时将功耗从18