ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

CUDA: use arch list for compatibility check (#11775)

Browse source 
* CUDA: use arch list for feature availability check

---------

Co-authored-by: Diego Devesa <slarengh@gmail.com>
This commit is contained in:
Johannes Gäßler 2025-02-11 00:17:22 +01:00 • committed by GitHub
parent 7b891bdc86
commit b9ab0a4d0b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 80 additions and 24 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/ggml-common.h
View file

@ -473,7 +473,6 @@ GGML_TABLE_BEGIN(uint8_t, ksigns_iq2xs, 128)
240, 113, 114, 243, 116, 245, 246, 119, 120, 249, 250, 123, 252, 125, 126, 255,
GGML_TABLE_END()
//#if __CUDA_ARCH__ >= GGML_CUDA_CC_DP4A // lowest compute capability for integer intrinsics
GGML_TABLE_BEGIN(uint64_t, ksigns64, 128)
0x0000000000000000, 0xff000000000000ff, 0xff0000000000ff00, 0x000000000000ffff,
0xff00000000ff0000, 0x0000000000ff00ff, 0x0000000000ffff00, 0xff00000000ffffff,
@ -508,7 +507,6 @@ GGML_TABLE_BEGIN(uint64_t, ksigns64, 128)
0x00ffffffff000000, 0xffffffffff0000ff, 0xffffffffff00ff00, 0x00ffffffff00ffff,
0xffffffffffff0000, 0x00ffffffffff00ff, 0x00ffffffffffff00, 0xffffffffffffffff,
GGML_TABLE_END()
//#endif
GGML_TABLE_BEGIN(uint64_t, iq2xxs_grid, 256)