注册并分享邀请链接,可获得视频播放与邀请奖励。

Aran Komatsuzaki
@arankomatsuzaki
Sharing AI research. Early work on AI (GPT-J, scaling, MoE). Ex ML PhD (GT) & Google.
加入 November 2016
437 正在关注    195.8K 粉丝
This feels like confusing a serving-runtime problem for a chip-startup opportunity. Agents do change inference patterns: loops, tool calls, branching, long context, KV reuse, burstiness. But most of that is an inference systems problem: scheduling, routing, KV-cache management, etc. Think Dynamo. By the time a new chip co tapes out + builds a compiler stack + wins cloud distribution, NVIDIA/AMD will likely have baked the obvious hardware-level optimizations into existing platforms.
显示更多
Inference Chips for Agent Workflows @sdianahu Most AI chips are designed for "prompt in, response out." Agents don't work that way. They loop, branch, and hold context across dozens of steps, and current GPUs hit 30–40% utilization as a result. That gap is where purpose-built silicon wins.
显示更多
0
14
99
10
转发到社区