single mention · inference & serving
DeepGEMM-based batch-invariant matrix multiplication
Replacing the existing matrix multiplication implementation end-to-end with DeepGEMM to support batch-invariant computation.
The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See inference kernel for the methods filed beside it.
Also called replace it end-to-end with DeepGEMM.
Filed under inference & serving › inference kernel