Skip to content

[WIP] Enable cooperative matrix support for Intel Arrow Lake H GPUs #18126

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
mmerecki wants to merge 1 commit into
base: master
Choose a base branch
from
+12 −3
Draft

[WIP] Enable cooperative matrix support for Intel Arrow Lake H GPUs #18126

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 12 additions & 3 deletions ggml/src/ggml-vulkan/ggml-vulkan.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2843,6 +2843,14 @@ static void ggml_vk_load_shaders(vk_device& device) {
m_warptile_mmq = m_warptile_mmq_int = { 256, 64, 64, 32, 16, 16, 2, 2, 2, 1, 16 };
m_warptile_mmqid = m_warptile_mmqid_int = { 256, 64, 64, 32, 16, 16, 2, 2, 2, 1, 16 };
}
#if defined(VK_KHR_cooperative_matrix) && defined(GGML_VULKAN_COOPMAT_GLSLC_SUPPORT)
else if ((device->vendor_id == VK_VENDOR_ID_INTEL) &&
(device->architecture != INTEL_XE2) &&
(device->driver_id == vk::DriverId::eIntelProprietaryWindows) &&
device->coopmat_support) {
m_warptile_mmq = { 256, 64, 64, 32, 16, 32, 2, tm_m, tn_m, tk_m, 32 };
}
#endif

l_mmq_wg_denoms = l_wg_denoms = {128, 128, 1 };
m_mmq_wg_denoms = m_wg_denoms = { 64, 64, 1 };
Expand Down Expand Up @@ -14446,9 +14454,10 @@ static bool ggml_vk_device_is_supported(const vk::PhysicalDevice & vkdev) {
static bool ggml_vk_khr_cooperative_matrix_support(const vk::PhysicalDeviceProperties& props, const vk::PhysicalDeviceDriverProperties& driver_props, vk_device_architecture arch) {
switch (props.vendorID) {
case VK_VENDOR_ID_INTEL:
// Only allowing Xe2 GPU at the moment since Xe2 GPU can gain significant performance boost,
// while some older hardware (ex. Arc A770) has performance regressions
return arch == vk_device_architecture::INTEL_XE2;
// Only allowing Xe2 GPU and integrated Xe GPUs at the moment since older hardware (ex. Arc A770) has performance regressions.
return arch == vk_device_architecture::INTEL_XE2 ||
(driver_props.driverID == vk::DriverId::eIntelProprietaryWindows &&
props.deviceType == vk::PhysicalDeviceType::eIntegratedGpu);
case VK_VENDOR_ID_AMD:
if (driver_props.driverID == vk::DriverId::eAmdProprietary || driver_props.driverID == vk::DriverId::eAmdOpenSource) {
// Workaround for AMD proprietary driver reporting support on all GPUs
Expand Down