云原生
jimmysong.io
A practitioner's perspective on AI infrastructure trends: evolving bottlenecks, roles of CPU/GPU/scheduling, ecosystem shifts, and compute demand across training, inference, and Agent workloads.…
云原生 的其他文章
- Models Keep Getting Bigger. Why Do We Still Need GPU Partitioning?
- KubeCon Returns to Shanghai - Cloud Native at the Crossroads of the AI Era
- Why GPU Scheduling Matters: From the Scheduling Landscape to Verifying HAMi mutex Semantics
- HAMi-core and KAI Scheduler: GPU Sharing Moves from Allocatable to Governable
- Terms of Service
讨论
登录后参与讨论
还没有评论，来说第一句吧。