Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
composable_kernel_ROCM
Commits
e7cde218
Commit
e7cde218
authored
Jul 03, 2024
by
Harisankar Sadasivan
Browse files
changes suggested in PR review are made- removing comments and correcting copyright
parent
57a38a1d
Changes
56
Show whitespace changes
Inline
Side-by-side
Showing
16 changed files
with
19 additions
and
41 deletions
+19
-41
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_default_instance.cpp
..._streamk_f16_f16_f16_mk_kn_mn_mem_v2_default_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_kpadding_instance.cpp
...streamk_f16_f16_f16_mk_kn_mn_mem_v2_kpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_mnkpadding_instance.cpp
...reamk_f16_f16_f16_mk_kn_mn_mem_v2_mnkpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp
...evice_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_default_instance.cpp
...al_streamk_f16_f16_f16_mk_nk_mn_comp_default_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_kpadding_instance.cpp
...l_streamk_f16_f16_f16_mk_nk_mn_comp_kpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_mnkpadding_instance.cpp
...streamk_f16_f16_f16_mk_nk_mn_comp_mnkpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_mnpadding_instance.cpp
..._streamk_f16_f16_f16_mk_nk_mn_comp_mnpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_default_instance.cpp
..._streamk_f16_f16_f16_mk_nk_mn_mem_v1_default_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_kpadding_instance.cpp
...streamk_f16_f16_f16_mk_nk_mn_mem_v1_kpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_mnkpadding_instance.cpp
...reamk_f16_f16_f16_mk_nk_mn_mem_v1_mnkpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_default_instance.cpp
..._streamk_f16_f16_f16_mk_nk_mn_mem_v2_default_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_kpadding_instance.cpp
...streamk_f16_f16_f16_mk_nk_mn_mem_v2_kpadding_instance.cpp
+1
-1
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_mnkpadding_instance.cpp
...reamk_f16_f16_f16_mk_nk_mn_mem_v2_mnkpadding_instance.cpp
+1
-1
profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp
.../include/profiler/profile_gemm_universal_streamk_impl.hpp
+4
-2
profiler/src/profile_gemm_universal_streamk.cpp
profiler/src/profile_gemm_universal_streamk.cpp
+1
-25
No files found.
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_default_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_kpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn_mem_v2_mnkpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_kn_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_default_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_kpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_mnkpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_comp_mnpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_default_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_kpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v1_mnkpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_default_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_kpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
library/src/tensor_operation_instance/gpu/gemm_universal_streamk/gemm_universal_streamk/device_gemm_xdl_universal_streamk_f16_f16_f16/device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn_mem_v2_mnkpadding_instance.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include "device_gemm_xdl_universal_streamk_f16_f16_f16_mk_nk_mn.hpp"
...
...
profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
...
...
@@ -153,7 +153,9 @@ bool profile_gemm_universal_streamk_impl(int do_verification,
for
(
auto
&
op_ptr
:
op_ptrs
)
{
std
::
vector
<
int
>
grid_size_list
=
{
38
,
76
,
114
,
152
,
190
,
228
,
266
,
304
,
342
,
380
};
std
::
vector
<
int
>
streamk_sel_list
=
{
0
,
1
,
2
,
3
,
4
};
std
::
vector
<
int
>
streamk_sel_list
=
{
0
,
1
,
2
,
3
,
4
};
// 0: Data Parallel (DP) mode (Stream-K OFF), 1: 1-tile Stream-K+ DP,
// 2:2-tile Stream-K + DP
if
(
Grid_size
==
-
1
)
{
...
...
profiler/src/profile_gemm_universal_streamk.cpp
View file @
e7cde218
// SPDX-License-Identifier: MIT
// Copyright (c) 20
18-2023
, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 20
24
, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream>
#include <numeric>
...
...
@@ -145,30 +145,6 @@ int profile_gemm_universal_streamk(int argc, char* argv[])
{
return
profile
(
F16
{},
F16
{},
F32
{},
F16
{},
Row
{},
Col
{},
Row
{});
}
// else if(data_type == GemmDataType::F16_F8_F16 && layout == GemmMatrixLayout::MK_KN_MN)
// {
// return profile(F16{}, F8{}, F32{}, F16{}, Row{}, Row{}, Row{});
// }
// else if(data_type == GemmDataType::F16_F8_F16 && layout == GemmMatrixLayout::MK_NK_MN)
// {
// return profile(F16{}, F8{}, F32{}, F16{}, Row{}, Col{}, Row{});
// }
// else if(data_type == GemmDataType::F8_F16_F16 && layout == GemmMatrixLayout::MK_KN_MN)
// {
// return profile(F8{}, F16{}, F32{}, F16{}, Row{}, Row{}, Row{});
// }
// else if(data_type == GemmDataType::F8_F16_F16 && layout == GemmMatrixLayout::MK_NK_MN)
// {
// return profile(F8{}, F16{}, F32{}, F16{}, Row{}, Col{}, Row{});
// }
// else if(data_type == GemmDataType::BF16_BF16_BF16 && layout == GemmMatrixLayout::MK_KN_MN)
// {
// return profile(BF16{}, BF16{}, F32{}, BF16{}, Row{}, Row{}, Row{});
// }
// else if(data_type == GemmDataType::BF16_BF16_BF16 && layout == GemmMatrixLayout::MK_NK_MN)
// {
// return profile(BF16{}, BF16{}, F32{}, BF16{}, Row{}, Col{}, Row{});
// }
else
{
std
::
cout
<<
"this data_type & layout is not implemented"
<<
std
::
endl
;
...
...
Prev
1
2
3
Next
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment