Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion sample_workloads/nccltest/docker/Dockerfile
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
FROM nvidia/cuda:12.0.0-devel-ubuntu20.04
FROM nvidia/cuda:12.2.2-devel-ubuntu20.04

ENV DEBIAN_FRONTEND='noninteractive'

Expand Down
7 changes: 4 additions & 3 deletions sample_workloads/nccltest/gke/values.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -27,12 +27,12 @@ telemetry:

rxdm:
image: "us-docker.pkg.dev/gce-ai-infra/gpudirect-tcpx/tcpgpudmarxd-dev"
tag: "v2.0.9"
tag: "v2.0.12"
flags: ["--setup_param \"--verbose 128 2 0\"", "--gpu_nic_preset a3vm", "--gpu_shmem_type fd"]

ncclPlugin:
image: "us-docker.pkg.dev/gce-ai-infra/gpudirect-tcpx/nccl-plugin-gpudirecttcpx-dev"
tag: "v3.1.6_2023_10_23"
tag: "v3.1.9-2.19.4-12.0"
unreservedCores: "0-7,104-111,52-59,156-163"
envs:
NCCL_GPUDIRECTTCPX_FORCE_ACK: "0"
Expand All @@ -51,9 +51,10 @@ ncclPlugin:
NCCL_MIN_NCHANNELS: 12
NCCL_GPUDIRECTTCPX_SOCKET_IFNAME: "eth1,eth2,eth3,eth4"
NCCL_GPUDIRECTTCPX_CTRL_DEV: "eth0"
NCCL_GPUDIRECTTCPX_PROGRAM_FLOW_STEERING_WAIT_MICROS: "1000000"
NCCL_GPUDIRECTTCPX_PROGRAM_FLOW_STEERING_WAIT_MICROS: "500000"
NCCL_CROSS_NIC: 0
NCCL_ALGO: "Ring"
NCCL_PROTO: "Simple"
NCCL_NET_GDR_LEVEL: "PIX"
NCCL_P2P_PXN_LEVEL: 0
NCCL_BUFFSIZE: 4194304