apply_mxfp8_scale#

nvmath.linalg.advanced.helpers.matmul.apply_mxfp8_scale(x, scale)[source]#

Apply MXFP8 block scale factors to tensor x.

Parameters:

x – The tensor to which the scaling should be applied.
scale – The block scale factors to apply.

Returns:

A float32 tensor with values of x with scales applied.

Note

This function is not intended for production usage due to its relatively low performance and high memory consumption. Instead of applying the scales manually using this function, use result_type to request non-FP8 output.