ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 21:21:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

[gfx110x] support Navi3x architectures. (#628)

Browse Source 
* enable building on Nav31

* fix syntax

* replace GPU_TARGETS with offload-arch

* add gfx1102 rachitecture

* fix typo

* update changelog
This commit is contained in:
Illia Silin
2023-03-09 05:56:40 -08:00
committed by GitHub
parent 9096b1c7b2
commit 0ccecc7c31
13 changed files with 27 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
include/ck/tensor_operation/gpu/device/impl/device_batched_contraction_multiple_d_wmma_cshuffle.hpp
View File

@@ -770,7 +770,8 @@ struct DeviceBatchedContractionMultipleD_Wmma_CShuffle
static bool IsSupportedArgument(const Argument& arg)
{
if(ck::get_device_name() == "gfx1100")
if(ck::get_device_name() == "gfx1100" || ck::get_device_name() == "gfx1101" ||
ck::get_device_name() == "gfx1102")
{
if constexpr(!(is_same_v<AccDataType, float> || is_same_v<AccDataType, int32_t>))
{

3
include/ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_wmma_cshuffle.hpp
View File

@@ -476,7 +476,8 @@ struct DeviceGemmMultipleD_Wmma_CShuffle : public DeviceGemmMultipleD<ALayout,
static bool IsSupportedArgument(const Argument& arg)
{
if(ck::get_device_name() == "gfx1100")
if(ck::get_device_name() == "gfx1100" || ck::get_device_name() == "gfx1101" ||
ck::get_device_name() == "gfx1102")
{
if constexpr(!(is_same_v<AccDataType, float> || is_same_v<AccDataType, int32_t>))
{

3
include/ck/tensor_operation/gpu/device/impl/device_gemm_wmma.hpp
View File

@@ -404,7 +404,8 @@ struct DeviceGemmWmma_CShuffle : public DeviceGemm<ALayout,
static bool IsSupportedArgument(const Argument& arg)
{
if(ck::get_device_name() == "gfx1100")
if(ck::get_device_name() == "gfx1100" || ck::get_device_name() == "gfx1101" ||
ck::get_device_name() == "gfx1102")
{
if constexpr(!(is_same_v<AccDataType, float> || is_same_v<AccDataType, int32_t>))
{

3
include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_wmma_cshuffle.hpp
View File

@@ -579,7 +579,8 @@ struct DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
namespace ctc = tensor_layout::convolution;
// check device
if(get_device_name() == "gfx1100")
if(get_device_name() == "gfx1100" || get_device_name() == "gfx1101" ||
ck::get_device_name() == "gfx1102")
{
if constexpr(!(is_same_v<AccDataType, float> || is_same_v<AccDataType, int32_t>))
{