LodeHQ
AI news for people who shipFree, by email

AI · github.com · Tuesday, 8 September 2026

VDN-H3 escapes the B200 cluster and lands in ComfyUI

A native ComfyUI node ports MiniMax-H3’s Video Delta Net to portable PyTorch, replacing datacenter-only FlashAttention and fused Triton kernels with consumer-GPU-compatible patches. The payoff is architectural: long-range video attention scales linearly with clip length, though this is still an experimental, VRAM-hungry route, not a way to reproduce the headline 74.5× speedup on an RTX 3090 Ti.

Read the source: github.com ↗In the 8 September issue →