MoonEP is open-sourced with a memory-saving MoE backprop design

SeunghyunSEO7 · x · 2026-07-28

MoonEP has been open-sourced, and the post highlights that it uses sonic backprop math to cut activation memory in MoE training.

The attached figure explains the memory-efficient permutation path: it reorders the router/permutation flow so fewer tensors need to be kept for backward pass, aiming to reduce activation memory while preserving the same functional computation. The author also notes the work is related to the NVIDIA Megatron team and credits Tri Dao’s usual influence on memory-saving design.

Related event: Moonshot Open-Sources MoonEP for MoE Expert Parallelism Communication(9 posts)→

Original post →

More from Infra

Infra channel →