Merge pull request #26 from coreweave/eta/nccl-2.18.5-and-cuda-12.2.2

build: Update to NCCL v2.18.5 & CUDA 12.2.2

.github/workflows/ubuntu-20.yml

@@ -48,7 +48,7 @@ jobs:
       base-tag: 12.0.1-cudnn8-devel-ubuntu20.04
       cuda-version-minor: "12.0.1"
       cuda-version-major: "12.0"
-      nccl-version: 2.18.3-1
+      nccl-version: 2.18.5-1
       cuda-samples-version: "12.0"
       hpcx-version: "2.16"
       hpcx-nccl-version: "2.18"
@@ -77,10 +77,10 @@ jobs:
       folder: .
       dockerfile: Dockerfile.ubuntu20
       base-image: nvidia/cuda
-      base-tag: 12.2.0-devel-ubuntu20.04
-      cuda-version-minor: "12.2.0"
+      base-tag: 12.2.2-cudnn8-devel-ubuntu20.04
+      cuda-version-minor: "12.2.2"
       cuda-version-major: "12.2"
-      nccl-version: 2.18.3-1
+      nccl-version: 2.18.5-1
       cuda-samples-version: "12.2"
       hpcx-version: "2.16"
       hpcx-nccl-version: "2.18"

Dockerfile.ubuntu20

@@ -20,7 +20,7 @@ RUN apt-get -qq update && \
 
 # Mellanox OFED (latest)
 RUN wget -qO - https://www.mellanox.com/downloads/ofed/RPM-GPG-KEY-Mellanox | apt-key add -
-RUN cd /etc/apt/sources.list.d/ && wget https://linux.mellanox.com/public/repo/mlnx_ofed/latest/ubuntu18.04/mellanox_mlnx_ofed.list
+RUN cd /etc/apt/sources.list.d/ && wget https://linux.mellanox.com/public/repo/mlnx_ofed/latest/ubuntu20.04/mellanox_mlnx_ofed.list
 
 RUN apt-get -qq update \
     && apt-get -qq install -y --no-install-recommends \

README.md

@@ -52,11 +52,11 @@ built from these Dockerfiles that can be used as base for your own images.
 
 | **Image Tag**                                                                     | **CUDA** | **NCCL** | **HPC-X** |
 |-----------------------------------------------------------------------------------|----------|----------|-----------|
-| ghcr.io/coreweave/nccl-tests:12.2.0-devel-ubuntu20.04-nccl2.18.3-1-253a5b1        | 12.2.0   | 2.18.3   | 2.16.0    |
+| ghcr.io/coreweave/nccl-tests:12.2.2-cudnn8-devel-ubuntu20.04-nccl2.18.5-1-a6a61ab | 12.2.2   | 2.18.5   | 2.16.0    |
 | ghcr.io/coreweave/nccl-tests:12.1.1-cudnn8-devel-ubuntu20.04-nccl2.18.3-1-253a5b1 | 12.1.1   | 2.18.3   | 2.16.0    |
-| ghcr.io/coreweave/nccl-tests:12.0.1-cudnn8-devel-ubuntu20.04-nccl2.18.3-1-253a5b1 | 12.0.1   | 2.18.3   | 2.16.0    |
-| ghcr.io/coreweave/nccl-tests:11.8.0-cudnn8-devel-ubuntu20.04-nccl2.16.2-1-253a5b1 | 11.8.0   | 2.16.2   | 2.14.0    |
-| ghcr.io/coreweave/nccl-tests:11.7.1-cudnn8-devel-ubuntu20.04-nccl2.14.3-1-253a5b1 | 11.7.1   | 2.14.3   | 2.14.0    |
+| ghcr.io/coreweave/nccl-tests:12.0.1-cudnn8-devel-ubuntu20.04-nccl2.18.5-1-a6a61ab | 12.0.1   | 2.18.5   | 2.16.0    |
+| ghcr.io/coreweave/nccl-tests:11.8.0-cudnn8-devel-ubuntu20.04-nccl2.16.2-1-a6a61ab | 11.8.0   | 2.16.2   | 2.14.0    |
+| ghcr.io/coreweave/nccl-tests:11.7.1-cudnn8-devel-ubuntu20.04-nccl2.14.3-1-a6a61ab | 11.7.1   | 2.14.3   | 2.14.0    |
 | coreweave/nccl-tests:2022-09-28_16-34-19.392_EDT                                  | 11.6.2   | 2.12.0   | 2.12      |
 
 ## Running NCCL Tests