2.7.3-1
Add support for A100 GPU and related platforms.
Add support for CUDA 11.
Add support for send/receive operations (beta).
[ROCm/rccl commit: 5949d96f36]
이 커밋은 다음에 포함됨:
1 부모
c43022b9d8
커밋
89072c82e5
54 파일이 변경됨
+2051
-1254
검색과 일치하는 파일 없음
@@ -1,5 +1,5 @@
|
||||
/*************************************************************************
|
||||
* Copyright (c) 2015-2019, NVIDIA CORPORATION. All rights reserved.
|
||||
* Copyright (c) 2015-2020, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* See LICENSE.txt for license information
|
||||
************************************************************************/
|
||||
@@ -50,19 +50,6 @@ static __inline__ int ncclTypeSize(ncclDataType_t type) {
|
||||
}
|
||||
}
|
||||
|
||||
#define NCCL_NUM_FUNCTIONS 5
|
||||
typedef enum { ncclCollBroadcast, ncclCollReduce, ncclCollAllGather, ncclCollReduceScatter, ncclCollAllReduce } ncclFunc_t;
|
||||
|
||||
#define NCCL_NUM_ALGORITHMS 3 // Tree/Ring/CollNet
|
||||
#define NCCL_ALGO_TREE 0
|
||||
#define NCCL_ALGO_RING 1
|
||||
#define NCCL_ALGO_COLLNET 2
|
||||
|
||||
#define NCCL_NUM_PROTOCOLS 3 // Simple/LL/LL128
|
||||
#define NCCL_PROTO_LL 0
|
||||
#define NCCL_PROTO_LL128 1
|
||||
#define NCCL_PROTO_SIMPLE 2
|
||||
|
||||
#include "debug.h"
|
||||
#include "checks.h"
|
||||
#include "alloc.h"
|
||||
|
||||
새 이슈에서 참조
사용자 차단