Merge remote-tracking branch 'nccl/master' into HEAD

src/graph/paths.cc

@@ -1,6 +1,6 @@
 /*************************************************************************
  * Copyright (c) 2018-2022, NVIDIA CORPORATION. All rights reserved.
- * Modifications Copyright (c) 2019-2022 Advanced Micro Devices, Inc. All rights reserved.
+ * Modifications Copyright (c) 2019-2023 Advanced Micro Devices, Inc. All rights reserved.
  *
  * See LICENSE.txt for license information
  ************************************************************************/
@@ -866,3 +866,15 @@ ncclResult_t ncclTopoGetNvbGpus(struct ncclTopoSystem* system, int rank, int* nr
   *nranks = nvbGpus;
   return ncclSuccess;
 }
+
+int ncclTopoPathAllNVLink(struct ncclTopoSystem* system) {
+  int minPath = PATH_DIS;
+  for (int i=0; i<system->nodes[GPU].count; i++) {
+    struct ncclTopoLinkList* paths = system->nodes[GPU].nodes[i].paths[GPU];
+    for (int j=0; j<system->nodes[GPU].count; j++) {
+      if (i == j) continue;
+      minPath = std::min(minPath, paths[j].type);
+    }
+  }
+  return minPath >= PATH_PIX ? 0 : 1;
+}

src/graph/topo.cc

@@ -1,6 +1,6 @@
 /*************************************************************************
  * Copyright (c) 2016-2022, NVIDIA CORPORATION. All rights reserved.
- * Modifications Copyright (c) 2019-2022 Advanced Micro Devices, Inc. All rights reserved.
+ * Modifications Copyright (c) 2019-2023 Advanced Micro Devices, Inc. All rights reserved.
  *
  * See LICENSE.txt for license information
  ************************************************************************/
@@ -78,6 +78,9 @@ static ncclResult_t ncclTopoGetInterCpuBw(struct ncclTopoNode* cpu, float* bw) {
   if (cpu->cpu.arch == NCCL_TOPO_CPU_ARCH_X86 && cpu->cpu.vendor == NCCL_TOPO_CPU_VENDOR_INTEL) {
     *bw = cpu->cpu.model == NCCL_TOPO_CPU_TYPE_SKL ? SKL_QPI_BW : QPI_BW;
   }
+  if (cpu->cpu.arch == NCCL_TOPO_CPU_ARCH_X86 && cpu->cpu.vendor == NCCL_TOPO_CPU_VENDOR_AMD) {
+    *bw = AMD_BW;
+  }
   if (cpu->cpu.arch == NCCL_TOPO_CPU_ARCH_X86 && cpu->cpu.vendor == NCCL_TOPO_CPU_VENDOR_ZHAOXIN) {
     *bw = cpu->cpu.model ==  NCCL_TOPO_CPU_TYPE_YONGFENG ? YONGFENG_ZPI_BW : ZPI_BW;
   }

src/graph/topo.h

@@ -1,6 +1,6 @@
 /*************************************************************************
  * Copyright (c) 2016-2022, NVIDIA CORPORATION. All rights reserved.
- * Modifications Copyright (c) 2019-2022 Advanced Micro Devices, Inc. All rights reserved.
+ * Modifications Copyright (c) 2019-2023 Advanced Micro Devices, Inc. All rights reserved.
  *
  * See LICENSE.txt for license information
  ************************************************************************/
@@ -19,6 +19,7 @@
 #define PCI_BW 12.0           // PCI Gen3 x16
 #define QPI_BW 6.0
 #define SKL_QPI_BW 9.0
+#define AMD_BW 16.0
 #define ZPI_BW 6.0
 #define YONGFENG_ZPI_BW 9.0
 #define P9_BW 32.0