Skip to content

New ComfyUI Optimizations for NVIDIA GPUs - NVFP4 Quantization, Async Offload, and Pinned Memory #482

Description

@naxci1

NVFP4 Quantization (Blackwell GPUs)
NVFP4 is a quantization format designed to make use of the FP4 hardware found on NVIDIA’s Blackwell architecture. When everything works correctly, RTX 50-series or Blackwell Pro GPU can get a ~2x performance boost compared to using fp8 or bf16/fp16 models.

An important caveat is that currently, ComfyUI only supports NVFP4 acceleration if you are running PyTorch built with CUDA 13.0 (cu130). Otherwise, while the model will still function, your sampling may actually be up to 2x slower than fp8. If you experience issues trying to get the full speed of NVFP4 models, checking your PyTorch version is the first thing you should try!

https://blog.comfy.org/p/new-comfyui-optimizations-for-nvidia

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions