Superminaren/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-27 21:46:57 +02:00
Code Issues Packages Projects Releases Wiki Activity

vulkan: tune KHR cooperative matrix support for Adreno GPUs (#29328)

Browse Source 
* Enable coopmat support for Vulkan backend

* Fixed the mul_mat_s

* Removed the debug statement
This commit is contained in:
Raman Shinde
2026-09-24 13:08:40 +02:00
committed by GitHub
parent 53ed051ce5
commit 3423f940e8
2 changed files with 28 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-vulkan/ggml-vulkan-types.h
+1
View File
@@ -390,6 +390,7 @@ enum vk_device_architecture {
INTEL_XE2,
NVIDIA_PRE_TURING,
NVIDIA_TURING,
QUALCOMM_ADRENO,
};
enum vk_conv_shapes {
ggml/src/ggml-vulkan/ggml-vulkan.cpp
+27 -2
View File
@@ -121,6 +121,23 @@ static vk_device_architecture get_device_architecture(const vk::PhysicalDevice&
return vk_device_architecture::NVIDIA_TURING;
}
}
} else if(props.vendorID == VK_VENDOR_ID_QUALCOMM){
const std::vector<vk::ExtensionProperties> ext_props = device.enumerateDeviceExtensionProperties();
bool cooperative_matrix = false;
bool cooperative_matrix_conversion = false;
for (const auto& properties : ext_props) {
if (strcmp("VK_KHR_cooperative_matrix", properties.extensionName) == 0) {
cooperative_matrix = true;
} else if (strcmp("VK_QCOM_cooperative_matrix_conversion", properties.extensionName) == 0) {
cooperative_matrix_conversion = true;
}
}
if (cooperative_matrix && cooperative_matrix_conversion) {
return vk_device_architecture::QUALCOMM_ADRENO;
}
}
return vk_device_architecture::OTHER;
}
@@ -1734,6 +1751,9 @@ void ggml_vk_load_shaders(vk_device& device, vk_pipeline requested) {
l_warptile = { 256, 128, 128, 16, mm_warp_8, 64, 2, tm_m, tn_m, tk_m, mm_warp_8 };
l_warptile_mmq = l_warptile_mmq_int = { 256, 128, 128, 32, mm_warp_8, 64, 2, tm_m, tn_m, tk_m, mm_warp_8 };
l_warptile_mmq_int_k = { 256, 128, 128, 32, mm_warp_16, 64, 1, 4, 2, 1, mm_warp_16 };
} else if (device->vendor_id == VK_VENDOR_ID_QUALCOMM && device->coopmat_support) {
m_warptile = { 64, 64, 64, 16, 64, 64, 1, tm_l, tn_l, tk_l, 64 };
m_warptile_mmq = { 64, 64, 64, 32, 64, 64, 1, tm_m, tn_m, tk_m, 64 };
}
l_mmq_wg_denoms = l_wg_denoms = {128, 128, 1 };
@@ -4606,10 +4626,10 @@ vk_device ggml_vk_get_device(size_t idx) {
case VK_VENDOR_ID_QUALCOMM:
device->mul_mat_l[i] = false;
device->mul_mat_m[i] = true;
device->mul_mat_s[i] = true;
device->mul_mat_s[i] = !device->coopmat_support;
device->mul_mat_id_l[i] = false;
device->mul_mat_id_m[i] = true;
device->mul_mat_id_s[i] = true;
device->mul_mat_id_s[i] = !device->coopmat_support;
break;
#endif
default:
@@ -6393,6 +6413,8 @@ static bool ggml_vk_should_use_mmvq(const vk_device& device, uint32_t m, uint32_
default:
return true;
}
case VK_VENDOR_ID_QUALCOMM:
return false;
default:
return true;
}
@@ -15956,6 +15978,9 @@ bool ggml_vk_khr_cooperative_matrix_support(const vk::PhysicalDeviceProperties&
return arch == vk_device_architecture::AMD_RDNA3;
}
return true;
case VK_VENDOR_ID_QUALCOMM:
// Only allow Adreno GPUs with hardware matrix cores (Gen 6+).
return arch == vk_device_architecture::QUALCOMM_ADRENO;
default:
return true;
}