From 261707d90a42255739fbabb30f0bd6fd88b1ffe1 Mon Sep 17 00:00:00 2001 From: Wenkai Du <43822138+wenkaidu@users.noreply.github.com> Date: Tue, 16 Jan 2024 07:54:18 -0800 Subject: [PATCH] Add option to force enable network transport on single node (#1046) --- src/graph/paths.cc | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/graph/paths.cc b/src/graph/paths.cc index 30de6a2d95..dbc5b2c3a1 100644 --- a/src/graph/paths.cc +++ b/src/graph/paths.cc @@ -771,6 +771,11 @@ ncclResult_t ncclTopoTrimSystem(struct ncclTopoSystem* system, struct ncclComm* INFO(NCCL_GRAPH, "GDR is available on all GPUs"); } + if (rcclParamEnableIntranet() == 1) { + remove = 0; + system->type |= RCCL_TOPO_FORCE_INTRA; + } + comm->localRanks = system->nodes[GPU].count; if (system->nodes[GPU].count == comm->nRanks && remove) { for (int n=system->nodes[NET].count-1; n>=0; n--)