Open Models Almanac
Methodsinference & servinginference quantization

single mention · inference & serving

NVFP4 weight-and-activation quantization

An NVFP4 recipe using 4-bit weights and activations on NVIDIA Blackwell.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See inference quantization for the methods filed beside it.

Also called NVFP4 recipe.

Filed under inference & serving › inference quantization