EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

vulkan: fuse adds (#15252)

Browse Source 
* vulkan: fuse adds

Fuse adds that have the same shape, which are common in MoE models.
It will currently fuse up to 6 adds, because we assume no more than
8 descriptors per dispatch. But this could be changed.

* check runtimeDescriptorArray feature

* disable multi_add for Intel due to likely driver bug
This commit is contained in:
Jeff Bolz
2025-08-16 11:48:22 -05:00
committed by GitHub
parent de2192794f
commit 1fe00296f5
6 changed files with 301 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
ggml/src/ggml-vulkan/vulkan-shaders/generic_binary_head.comp
View File

@@ -2,6 +2,7 @@
#extension GL_EXT_control_flow_attributes : require
#include "rte.comp"
#include "utils.comp"
layout (push_constant) uniform parameter
{
@@ -28,25 +29,9 @@ uint get_aoffset() { return p.misalign_offsets >> 16; }
uint get_boffset() { return (p.misalign_offsets >> 8) & 0xFF; }
uint get_doffset() { return p.misalign_offsets & 0xFF; }
// mod and div are expensive and coordinates/dimensions are often power of 2 or equal to 1
uint fastmod(uint a, uint b) {
if ((b & (b-1)) == 0) {
return a & (b-1);
}
return a % b;
}
uint fastdiv(uint a, uint b) {
return (a < b) ? 0 : (a / b);
}
void get_indices(uint idx, out uint i00, out uint i01, out uint i02, out uint i03) {
i03 = fastdiv(idx, (p.ne02*p.ne01*p.ne00));
const uint i03_offset = i03 * p.ne02*p.ne01*p.ne00;
i02 = fastdiv((idx - i03_offset), (p.ne01*p.ne00));
const uint i02_offset = i02*p.ne01*p.ne00;
i01 = (idx - i03_offset - i02_offset) / p.ne00;
i00 = idx - i03_offset - i02_offset - i01*p.ne00;
get_indices(idx, i00, i01, i02, i03, p.ne00, p.ne01, p.ne02, p.ne03);
}
uint src0_idx(uint i00, uint i01, uint i02, uint i03) {