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
OpenDAS
dgl
Commits
3befaca2
Commit
3befaca2
authored
Oct 16, 2024
by
sangwzh
Browse files
update warpsize to 64
parent
910cec0c
Changes
5
Hide whitespace changes
Inline
Side-by-side
Showing
5 changed files
with
7 additions
and
6 deletions
+7
-6
graphbolt/src/cuda/gpu_cache.h
graphbolt/src/cuda/gpu_cache.h
+1
-1
src/array/cuda/array_index_select.hip
src/array/cuda/array_index_select.hip
+2
-1
src/runtime/cuda/gpu_cache.hip
src/runtime/cuda/gpu_cache.hip
+1
-1
third_party/HugeCTR/gpu_cache/include/nv_gpu_cache.hpp
third_party/HugeCTR/gpu_cache/include/nv_gpu_cache.hpp
+1
-1
third_party/HugeCTR/gpu_cache/src/nv_gpu_cache.hip
third_party/HugeCTR/gpu_cache/src/nv_gpu_cache.hip
+2
-2
No files found.
graphbolt/src/cuda/gpu_cache.h
View file @
3befaca2
...
...
@@ -20,7 +20,7 @@ namespace cuda {
class
GpuCache
:
public
torch
::
CustomClassHolder
{
using
key_t
=
long
long
;
constexpr
static
int
set_associativity
=
2
;
constexpr
static
int
WARP_SIZE
=
32
;
constexpr
static
int
WARP_SIZE
=
64
;
constexpr
static
int
bucket_size
=
WARP_SIZE
*
set_associativity
;
using
gpu_cache_t
=
::
gpu_cache
::
gpu_cache
<
key_t
,
uint64_t
,
std
::
numeric_limits
<
key_t
>::
max
(),
set_associativity
,
...
...
src/array/cuda/array_index_select.hip
View file @
3befaca2
...
...
@@ -34,7 +34,8 @@ NDArray IndexSelect(NDArray array, IdArray index) {
DType* ret_data = static_cast<DType*>(ret->data);
const DType* array_data = static_cast<DType*>(cuda::GetDevicePointer(array));
const IdType* idx_data = static_cast<IdType*>(index->data);
// const IdType* idx_data = static_cast<IdType*>(index->data);
const IdType* idx_data = static_cast<IdType*>(cuda::GetDevicePointer(index));
hipStream_t stream = runtime::getCurrentHIPStreamMasqueradingAsCUDA();
if (num_feat == 1) {
...
...
src/runtime/cuda/gpu_cache.hip
View file @
3befaca2
...
...
@@ -41,7 +41,7 @@ namespace cuda {
template <typename key_t>
class GpuCache : public runtime::Object {
constexpr static int set_associativity = 2;
constexpr static int WARP_SIZE =
32
;
constexpr static int WARP_SIZE =
64
;
constexpr static int bucket_size = WARP_SIZE * set_associativity;
using gpu_cache_t = gpu_cache::gpu_cache<
key_t, uint64_t, std::numeric_limits<key_t>::max(), set_associativity,
...
...
third_party/HugeCTR/gpu_cache/include/nv_gpu_cache.hpp
View file @
3befaca2
...
...
@@ -30,7 +30,7 @@
#endif
#define SET_ASSOCIATIVITY 2
#define SLAB_SIZE
32
#define SLAB_SIZE
64
#define TASK_PER_WARP_TILE_MACRO 1
namespace
gpu_cache
{
...
...
third_party/HugeCTR/gpu_cache/src/nv_gpu_cache.hip
View file @
3befaca2
...
...
@@ -1251,7 +1251,7 @@ gpu_cache<key_type, ref_counter_type, empty_key, set_associativity, warp_size, s
return;
}
if (warp_size != 1 && warp_size != 2 && warp_size != 4 && warp_size != 8 && warp_size != 16 &&
warp_size != 32) {
warp_size != 32
&&warp_size != 64
) {
printf("Error: Invalid value for warp_size.\n");
return;
}
...
...
@@ -1299,7 +1299,7 @@ gpu_cache<key_type, ref_counter_type, empty_key, set_associativity, warp_size, s
return;
}
if (warp_size != 1 && warp_size != 2 && warp_size != 4 && warp_size != 8 && warp_size != 16 &&
warp_size != 32) {
warp_size != 32
&& warp_size != 64
) {
printf("Error: Invalid value for warp_size.\n");
return;
}
...
...
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