search

docker / for-win

Bug reports for Docker Desktop for Windows
https://www.docker.com/products/docker#/windows
1.87k stars 289 forks source link

Windows 10, NVidia GPU works in Docker Engine, but it doesn't in Docker Desktop app #14015

Open EntityinArray opened 7 months ago

EntityinArray commented 7 months ago

Description

Executing this command in Docker Engine works fine, NVidia RTX 3060 is detected:

PS C:\Users\entityinarray> docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark
Unable to find image 'nvcr.io/nvidia/k8s/cuda-sample:nbody' locally
nbody: Pulling from nvidia/k8s/cuda-sample
22c5ef60a68e: Pull complete
1939e4248814: Pull complete
548afb82c856: Pull complete
a424d45fd86f: Pull complete
207b64ab7ce6: Pull complete
f65423f1b49b: Pull complete
2b60900a3ea5: Pull complete
e9bff09d04df: Pull complete
edc14edf1b04: Pull complete
1f37f461c076: Pull complete
9026fb14bf88: Pull complete
Digest: sha256:59261e419d6d48a772aad5bb213f9f1588fcdb042b115ceb7166c89a51f03363
Status: Downloaded newer image for nvcr.io/nvidia/k8s/cuda-sample:nbody
Run "nbody -benchmark [-numbodies=<numBodies>]" to measure performance.
        -fullscreen       (run n-body simulation in fullscreen mode)
        -fp64             (use double precision floating point values for simulation)
        -hostmem          (stores simulation data in host memory)
        -benchmark        (run benchmark to measure performance)
        -numbodies=<N>    (number of bodies (>= 1) to run in simulation)
        -device=<d>       (where d=0,1,2.... for the CUDA device to use)
        -numdevices=<i>   (where i=(number of CUDA devices > 0) to use for simulation)
        -compare          (compares simulation results running once on the default GPU and once on the CPU)
        -cpu              (run n-body simulation on the CPU)
        -tipsy=<file.bin> (load a tipsy model file for simulation)

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

> Windowed mode
> Simulation data stored in video memory
> Single precision floating point simulation
> 1 Devices used for simulation
GPU Device 0: "Ampere" with compute capability 8.6

> Compute 8.6 CUDA device: [NVIDIA GeForce RTX 3060]
28672 bodies, total time for 10 iterations: 23.076 ms
= 356.253 billion interactions per second
= 7125.059 single-precision GFLOP/s at 20 flops per interaction

However, using any container via Docker Desktop app fails to expose my NVidia card: image

Reproduce

  1. Execute docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark and have NVidia card detected.
  2. Try to download and run any container via Docker Desktop App (NOT Docker Engine CLI command)
  3. It doesn't detect your GPU

Expected behavior

GPU should be detected

docker version

Client:
 Cloud integration: v1.0.35+desktop.13
 Version:           26.0.0
 API version:       1.45
 Go version:        go1.21.8
 Git commit:        2ae903e
 Built:             Wed Mar 20 15:18:56 2024
 OS/Arch:           windows/amd64
 Context:           default

Server: Docker Desktop 4.29.0 (145265)
 Engine:
  Version:          26.0.0
  API version:      1.45 (minimum version 1.24)
  Go version:       go1.21.8
  Git commit:       8b79278
  Built:            Wed Mar 20 15:18:01 2024
  OS/Arch:          linux/amd64
  Experimental:     false
 containerd:
  Version:          1.6.28
  GitCommit:        ae07eda36dd25f8a1b98dfbf587313b99c0190bb
 runc:
  Version:          1.1.12
  GitCommit:        v1.1.12-0-g51d5e94
 docker-init:
  Version:          0.19.0
  GitCommit:        de40ad0

docker info

Client:
 Version:    26.0.0
 Context:    default
 Debug Mode: false
 Plugins:
  buildx: Docker Buildx (Docker Inc.)
    Version:  v0.13.1-desktop.1
    Path:     C:\Program Files\Docker\cli-plugins\docker-buildx.exe
  compose: Docker Compose (Docker Inc.)
    Version:  v2.26.1-desktop.1
    Path:     C:\Program Files\Docker\cli-plugins\docker-compose.exe
  debug: Get a shell into any image or container. (Docker Inc.)
    Version:  0.0.27
    Path:     C:\Program Files\Docker\cli-plugins\docker-debug.exe
  dev: Docker Dev Environments (Docker Inc.)
    Version:  v0.1.2
    Path:     C:\Program Files\Docker\cli-plugins\docker-dev.exe
  extension: Manages Docker extensions (Docker Inc.)
    Version:  v0.2.23
    Path:     C:\Program Files\Docker\cli-plugins\docker-extension.exe
  feedback: Provide feedback, right in your terminal! (Docker Inc.)
    Version:  v1.0.4
    Path:     C:\Program Files\Docker\cli-plugins\docker-feedback.exe
  init: Creates Docker-related starter files for your project (Docker Inc.)
    Version:  v1.1.0
    Path:     C:\Program Files\Docker\cli-plugins\docker-init.exe
  sbom: View the packaged-based Software Bill Of Materials (SBOM) for an image (Anchore Inc.)
    Version:  0.6.0
    Path:     C:\Program Files\Docker\cli-plugins\docker-sbom.exe
  scout: Docker Scout (Docker Inc.)
    Version:  v1.6.3
    Path:     C:\Program Files\Docker\cli-plugins\docker-scout.exe

Server:
 Containers: 2
  Running: 0
  Paused: 0
  Stopped: 2
 Images: 2
 Server Version: 26.0.0
 Storage Driver: overlay2
  Backing Filesystem: extfs
  Supports d_type: true
  Using metacopy: false
  Native Overlay Diff: true
  userxattr: false
 Logging Driver: json-file
 Cgroup Driver: cgroupfs
 Cgroup Version: 1
 Plugins:
  Volume: local
  Network: bridge host ipvlan macvlan null overlay
  Log: awslogs fluentd gcplogs gelf journald json-file local splunk syslog
 Swarm: inactive
 Runtimes: runc io.containerd.runc.v2
 Default Runtime: runc
 Init Binary: docker-init
 containerd version: ae07eda36dd25f8a1b98dfbf587313b99c0190bb
 runc version: v1.1.12-0-g51d5e94
 init version: de40ad0
 Security Options:
  seccomp
   Profile: unconfined
 Kernel Version: 5.15.146.1-microsoft-standard-WSL2
 Operating System: Docker Desktop
 OSType: linux
 Architecture: x86_64
 CPUs: 16
 Total Memory: 15.57GiB
 Name: docker-desktop
 ID: 10c3f722-bde1-455d-a192-29c0e64eac94
 Docker Root Dir: /var/lib/docker
 Debug Mode: false
 HTTP Proxy: http.docker.internal:3128
 HTTPS Proxy: http.docker.internal:3128
 No Proxy: hubproxy.docker.internal
 Labels:
  com.docker.desktop.address=npipe://\\.\pipe\docker_cli
 Experimental: false
 Insecure Registries:
  hubproxy.docker.internal:5555
  127.0.0.0/8
 Live Restore Enabled: false

WARNING: No blkio throttle.read_bps_device support
WARNING: No blkio throttle.write_bps_device support
WARNING: No blkio throttle.read_iops_device support
WARNING: No blkio throttle.write_iops_device support
WARNING: daemon is not using the default seccomp profile

Diagnostics ID

8E51C7E1-4513-4ACD-AE1B-DEBC2A042F90/20240413082627

Additional Info

Every Google search for this problem leads to someone just reccomending to do docker run --gpus all, but that's not what's needed.

MihaelaStoica commented 7 months ago

@EntityinArray, starting the container with the --gpus flag is exactly what is needed to expose the GPUs to the container. Starting the container directly from the Docker Desktop GUI doesn't add this flag, so the container cannot access the GPU resources.

EntityinArray commented 7 months ago

@EntityinArray, starting the container with the --gpus flag is exactly what is needed to expose the GPUs to the container. Starting the container directly from the Docker Desktop GUI doesn't add this flag, so the container cannot access the GPU resources.

Cool, how to add it in GUI?

I4m5umm3r commented 2 months ago

@EntityinArray, starting the container with the --gpus flag is exactly what is needed to expose the GPUs to the container. Starting the container directly from the Docker Desktop GUI doesn't add this flag, so the container cannot access the GPU resources.

Cool, how to add it in GUI?

did this ever get answered, or solved? I appear to have the same problem. 👍

edit: this is the most useful post I've found on this subject on the whole dear www

EntityinArray commented 2 months ago

Currently, no. Looks like the only way to assign GPUs to a container is on the command line. docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark docker run --gpus all your-container-name

I4m5umm3r commented 2 months ago

cli? :]

I4m5umm3r commented 2 months ago

ah. doh. I'll see if I can break everything by trying this out...not that I have a clue what I'm doing (arty type) big cheers for the post and response 🥇

I4m5umm3r commented 2 months ago

Compute 8.9 CUDA device: [NVIDIA GeForce RTX 4060 Laptop GPU] 24576 bodies, total time for 10 iterations: 17.717 ms = 340.899 billion interactions per second = 6817.986 single-precision GFLOP/s at 20 flops per interaction Unable to find image 'fooocus-main:latest' locally docker: Error response from daemon: pull access denied for fooocus-main, repository does not exist or may require 'docker login'. See 'docker run --help'.

container is fooocus-main

bit confused, which is my natural state, but getting a headache now...don't fell you need to reply, I'll keep at it, slooowly.

joyjoy

ps, fyi https://forums.docker.com/t/nvidia-cuda-doesnt-work-on-docker-desktop-but-works-on-docker-engine/130668/5