筛选结果
7 篇排序
1:04:26EN → 简体中文
Sequoia Capital· 2026年10月6日播客100,000 个加速器，每小时都有东西出故障Google AI 基础设施负责人 Amin Vahdat 向红杉资本的 Sonya Huang 解释，为什么 FLOPS 是虚荣指标，为什么 TPU 产品线一分为二，以及为什么他宁愿等电网接入，也不自己建一座吉瓦级的数据中心。
1:10:15EN → 简体中文Sequoia Capital· 2026年6月30日播客100 倍藏在各层之间：Dylan Patel 谈协同设计，以及一年降本 60 倍SemiAnalysis 创始人认为，让模型、算子和芯片彼此匹配，才能把这里的 2 倍、那里的 2 倍叠加成 100 倍；与此同时，同等质量的推理（inference）成本一年里便宜了约 60 倍。
44:19ENSequoia Capital· 2026年3月17日播客The $500-per-Kilogram Line: Why Starcloud Thinks Data Centers Belong in SpaceStarcloud founder Philip Johnston walks through the physics, launch economics, and business model of putting AI compute in orbit.
1:02:19ENSequoia Capital· 2025年12月10日播客A Top Video Model's Half-Life Is 30 Days. fal Runs 600 AnywayThe fal founders tell Sequoia why generative video is a compute-bound problem that language-model shops keep underestimating, and why they will not let any customer marry a single model.
41:32ENSequoia Capital· 2025年10月28日播客Moore's Law Is Slowing. Nvidia's CTO Still Wants 10× a YearMichael Kagan on why Nvidia's $7 billion Mellanox deal, not just more transistors, is how you turn a data center into one computer.
39:18ENSequoia Capital· 2024年8月13日播客Open and Closed Will Tie at the Same Size: PyTorch's Former Lead on Fast InferenceA six-month library swap at Meta became a five-year rebuild. Lin Qiao wants the rest of the industry to get that stack in five weeks.
1:00:28ENSequoia Capital· 2024年7月9日播客Scaling Laws Aren't Done. Don't Lock Yourself Into Yesterday's Architecture.Microsoft CTO Kevin Scott on why the exponential is still compounding, why inference will dwarf training, and why a language model is the world's crappiest database.