opensource/developer-roadmap
3
0
Fork 0
mirror of https://github.com/kamranahmedse/developer-roadmap.git synced 2026-10-25 09:34:00 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
remove-feed-resources
developer-roadmap/roadmaps/inference-engineering/content/expert-parallelism@62y78y-RXZT_QCUfd-YFE.md
T

655 B
Raw Permalink Blame History

Expert Parallelism

Expert Parallelism distributes MoE experts across GPUs so each GPU hosts multiple complete experts. Tokens are routed to the GPU holding their activated experts, and each expert runs independently within a single GPU without splitting across multiple. EP has lower inter-GPU communication overhead than Tensor Parallelism and scales well to multi-node deployments via InfiniBand.

Visit the following resources to learn more:

Reference in New Issue View Git Blame Copy Permalink