
Infrastructure & AI | Amin Vahdat, Google & Jeetu Patel
基础设施与 AI | Google 的 Amin Vahdat 对谈 Jeetu Patel
39 分钟阅读Podcasts
【技术前瞻】谷歌基建掌门人:AI 时代的算力竞赛与数据中心进化论 当模型规模指数级增长,支撑它们的硬件基建正面临怎样的挑战?谷歌基础设施负责人 Amin Vahdat 详细拆解了张量处理器(TPU)、液冷技术以及全球网络延迟优化背后的工程逻辑。这是了解 AI 下半场“物理层”竞争的权威指南。 🎙️ 访谈人物介绍 Amin Vahdat:谷歌系统基础设施副总裁、院士。他是计算机网络和分布式系统领域的顶级专家。 核心议题:TPU 的演进、数据中心能源挑战、模型推理的带宽瓶颈、谷歌算力版图。 关键词:谷歌基建、TPU、Amin Vahdat、数据中心、算力
章节 1:Google的逆袭心态与Gemini 3的里程碑
📝 本节摘要:
本节对话作为开场,聚焦于Google最新发布的模型 Gemini 3。嘉宾Amin Vahdat分享了Google在过去三年AI竞赛中的心路历程,坦言Google曾一度被视为“处于劣势的一方”(Underdog)。他描述了内部团队如何在外界质疑声中坚持研发,最终通过Gemini 3在各项基准测试中重回巅峰,成功证明了唱衰者(Naysayers)的错误。
[原文] [Jeetu Patel]: how's life life is exciting i mean it's actually never been more exciting i think I know you feel it i know you all feel it in the audience uh I'm I'm just thrilled actually Gemini 3 is kind of uh crazy times at Google huh
[译文] [Jeetu Patel]: 生活怎么样?生活很令人兴奋,我是说,实际上从未像现在这样令人兴奋。我想我知道你有这种感觉,我知道观众席上的各位都有这种感觉。呃,我真的非常激动。Gemini 3(的发布)... Google 现在的日子有点疯狂,是吧?
[原文] [Amin Vahdat]: gemini 3 has been awesome i mean I think it's been uh you all know it's uh uh state-of-the-art across essentially all the benchmarks we're really proud of that but I think for us you know we've been on this three plus year journey at Google gemini one came out twoish years ago uh I mean I think that for me I'll just speak for myself i'm I'm always rooting for the underdog and it was great to be a place where Google was the underdog you're always rooting for the underdog rooting for the underdog it was great to be at a place where Google was the underdog all of us believed internally actually all of us believed internally but for us to uh come to this journey the race is not I mean the race is is like any one right but uh we're proud of where we've come over the past few years
[译文] [Amin Vahdat]: Gemini 3 非常棒。我的意思是,我想大家都知道,它在几乎所有的基准测试(benchmarks)中都处于最先进水平(state-of-the-art),我们对此感到非常自豪。但在我看来,对于我们来说,你知道我们在 Google 已经走过了三年多的旅程。Gemini 1 大约是两年前发布的。呃,我是说,就我个人而言,我只代表我自己说,我总是支持处于劣势的一方(underdog)。能在 Google 处于劣势的时候身在其中,感觉很棒。你总是支持劣势方,支持劣势方。能在 Google 被视为劣势方的时候待在这里感觉很棒。我们在内部都坚信,实际上我们所有人在内部都坚信这一点。但对于我们要走完这段旅程... 这场竞赛并不是... 我的意思是这场竞赛就像任何其他竞赛一样,对吧?但呃,我们对过去几年取得的成就感到自豪。
[原文] [Jeetu Patel]: you should you should be very proud it was actually um it's it by the way you proved the naysayers wrong in some pretty profound ways because people had almost given up on Google in certain areas and then all of a sudden they're they're back and so it's a wild success and congratulations to you couldn't have better people kind of group of people
[译文] [Jeetu Patel]: 你们应该,你们应该感到非常自豪。实际上,呃... 顺便说一句,你们以一种非常深刻的方式证明了唱反调的人(naysayers)是错的。因为在某些领域,人们几乎已经放弃了 Google,然后突然之间,他们(Google)又回来了。所以这是一个巨大的成功(wild success),祝贺你们。不可能有比这更好的人员... 这样的一群人了。
- 章节 2:全栈协同——Google的“秘密武器”
- 章节 3:专用芯片(XPU)时代的算力哲学
- 章节 4:挑战物理极限——缩短硬件迭代周期
- 章节 5:终极前沿——太空数据中心构想
- 章节 6:增长的焦虑——速度、能源与效率悖论
- 章节 7:AI的真正价值——从“原创性”到“全人类的导师”
- 章节 8:摩尔定律的回归与未来展望
正在加载后续章节…







