[Build] Upgrade runtime image to Ubuntu 24.04, pick up rdma-core > 44 (#51058)

Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>

docker/Dockerfile

@@ -24,7 +24,7 @@
 
 ARG CUDA_VERSION=13.0.3
 ARG PYTHON_VERSION=3.12
-ARG UBUNTU_VERSION=22.04
+ARG UBUNTU_VERSION=24.04
 # DeepEPv2 requires NCCL >= 2.30.4 (GIN backend).
 # This version is used for CUDA 12+ builds to keep the DeepEP build and
 # runtime NCCL versions compatible.
@@ -765,7 +765,7 @@ RUN --mount=type=cache,target=/opt/uv/cache \
 # Install gdrcopy (saves ~6s per build)
 # TODO (huydhn): There is no prebuilt gdrcopy package on 12.9 at the moment
 ARG GDRCOPY_CUDA_VERSION=12.8
-ARG GDRCOPY_OS_VERSION=Ubuntu22_04
+ARG GDRCOPY_OS_VERSION=Ubuntu24_04
 ARG TARGETPLATFORM
 COPY tools/install_gdrcopy.sh /tmp/install_gdrcopy.sh
 RUN set -eux; \

docker/versions.json

@@ -8,7 +8,7 @@
       "default": "3.12"
     },
     "UBUNTU_VERSION": {
-      "default": "22.04"
+      "default": "24.04"
     },
     "NCCL_VERSION": {
       "default": "2.30.7"
@@ -17,7 +17,7 @@
       "default": "pytorch/manylinux2_28-builder:cuda13.0"
     },
     "FINAL_BASE_IMAGE": {
-      "default": "nvidia/cuda:13.0.3-base-ubuntu22.04"
+      "default": "nvidia/cuda:13.0.3-base-ubuntu24.04"
     },
     "GET_PIP_URL": {
       "default": "https://bootstrap.pypa.io/get-pip.py"
@@ -74,7 +74,7 @@
       "default": "12.8"
     },
     "GDRCOPY_OS_VERSION": {
-      "default": "Ubuntu22_04"
+      "default": "Ubuntu24_04"
     },
     "BITSANDBYTES_VERSION_X86": {
       "default": "0.46.1"

docs/getting_started/installation/gpu.cuda.inc.md

@@ -321,6 +321,8 @@ You can add any other [engine-args](https://docs.vllm.ai/en/latest/configuration
 
 vLLM's Docker image comes with [CUDA compatibility libraries](https://docs.nvidia.com/deploy/cuda-compatibility/index.html) pre-installed. This allows you to run vLLM on systems with NVIDIA drivers that are older than the CUDA Toolkit version used in the image, but only supports select professional and datacenter NVIDIA GPUs.
 
+For CUDA 13 images, the minimum host kernel is Linux 4.15 when running normally because CUDA 13 requires an R580 or newer driver. Compatibility mode supports R535 and R570 host drivers; R535 lowers the minimum host kernel to Linux 3.10, while R570 still requires Linux 4.15. Upgrading the container userland to Ubuntu 24.04 does not raise these driver-defined kernel requirements. See NVIDIA's [forward compatibility matrix](https://docs.nvidia.com/deploy/cuda-compatibility/forward-compatibility.html#use-the-right-cuda-forward-compatibility-package), [R535 requirements](https://download.nvidia.com/XFree86/Linux-x86_64/535.104.05/README/minimumrequirements.html), and [R580 requirements](https://download.nvidia.com/XFree86/Linux-x86_64/580.76.05/README/minimumrequirements.html).
+
 To enable this feature, set the `VLLM_ENABLE_CUDA_COMPATIBILITY` environment variable to `1` or `true` when running the container:
 
 ```bash