vulkan: Optimize contiguous copies (#10254)
* tests: Fix memory bandwidth calculation for perf tests Add a flops calculation for flash attention. Add one GGML_OP_CPY perf test. * vulkan: Optimize contiguous copies Add a variant of the copy shader for when the tensors are contiguous. Avoid the complex addressing calculations, and do four elements per invocation to hide some other overhead. Apply similar changes to the scale shader, since scale is always contiguous. Add a "progress bar" for shader compiles.
This commit is contained in:
parent
54ef9cfc72
commit
80dd7ff22f
13 changed files with 144 additions and 27 deletions
|
@ -350,6 +350,9 @@ void process_shaders() {
|
|||
string_to_spv("cpy_f32_f32", "copy.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}});
|
||||
string_to_spv("cpy_f32_f16", "copy.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float16_t"}});
|
||||
string_to_spv("cpy_f16_f16", "copy.comp", {{"A_TYPE", "float16_t"}, {"D_TYPE", "float16_t"}, {"OPTIMIZATION_ERROR_WORKAROUND", "1"}});
|
||||
string_to_spv("contig_cpy_f32_f32", "contig_copy.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float"}});
|
||||
string_to_spv("contig_cpy_f32_f16", "contig_copy.comp", {{"A_TYPE", "float"}, {"D_TYPE", "float16_t"}});
|
||||
string_to_spv("contig_cpy_f16_f16", "contig_copy.comp", {{"A_TYPE", "float16_t"}, {"D_TYPE", "float16_t"}, {"OPTIMIZATION_ERROR_WORKAROUND", "1"}});
|
||||
|
||||
string_to_spv("add_f32", "add.comp", {{"A_TYPE", "float"}, {"B_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});
|
||||
string_to_spv("add_f16_f32_f16", "add.comp", {{"A_TYPE", "float16_t"}, {"B_TYPE", "float"}, {"D_TYPE", "float16_t"}, {"FLOAT_TYPE", "float"}});
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue