vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-10-03 03:25:40 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

cuda : tune MMVQ to MMQ crossover for SM70 (Volta) (#28912)
Some checks failed
Python Type-Check / python type-check (push) Has been cancelled
Details
Update Operations Documentation / update-ops-docs (push) Has been cancelled
Details

Browse source 
* tune MMVQ to MMQ crossover for SM70 (Volta)

Signed-off-by: Yangyu Chen <cyy@cyyself.name>

* Apply suggestion from @JohannesGaessler

* Apply suggestion from @JohannesGaessler

* Apply suggestion from @JohannesGaessler

---------

Signed-off-by: Yangyu Chen <cyy@cyyself.name>
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
Yangyu Chen 2026-09-21 15:45:31 +08:00 • committed by GitHub
parent 8aa161b54a
commit 68d9053afd
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 16 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

16
ggml/src/ggml-cuda/mmvq.cu
View file

@ -365,6 +365,22 @@ bool ggml_cuda_should_use_mmvq(enum ggml_type type, int cc, int64_t ne11) {
return ne11 <= MMVQ_MAX_BATCH_SIZE;
}
}
if (GGML_CUDA_CC_IS_NVIDIA(cc) && cc == GGML_CUDA_CC_VOLTA) {
switch (type) {
case GGML_TYPE_Q2_K:
return ne11 <= 4;
case GGML_TYPE_Q3_K:
return ne11 <= 6;
case GGML_TYPE_Q4_K:
return ne11 <= 5;
case GGML_TYPE_Q5_K:
return ne11 <= 6;
case GGML_TYPE_Q6_K:
return ne11 <= 7;
default:
return ne11 <= MMVQ_MAX_BATCH_SIZE;
}
}
if (GGML_CUDA_CC_IS_CDNA(cc)) {
if (GGML_CUDA_CC_IS_CDNA1(cc)) {
switch (type) {