gpu-programming-101/docker/docker-compose.yml at e3b6f95889af6328e668af358e1234df6be89fc7 · AIComputing101/gpu-programming-101 · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
# GPU Programming 101 - Docker Compose Configuration
# Supports both NVIDIA CUDA and AMD ROCm platforms
# Updated for CUDA 12.9.1 and ROCm 7.0 (2025)

services:
  # NVIDIA CUDA Development Environment
  cuda-dev:
    build:
      context: ../
      dockerfile: docker/cuda/Dockerfile
    image: gpu-programming-101:cuda
    container_name: gpu101-cuda-dev
    hostname: cuda-dev

    # GPU access for NVIDIA
    runtime: nvidia
    environment:
      - NVIDIA_VISIBLE_DEVICES=all
      - NVIDIA_DRIVER_CAPABILITIES=compute,utility,graphics

    # Volume mounts
    volumes:
      - ../:/workspace/gpu-programming-101:rw
      - cuda-home:/root
      - cuda-cache:/root/.cache

    # Interactive mode
    stdin_open: true
    tty: true

    # Working directory
    working_dir: /workspace/gpu-programming-101

    # Network
    networks:
      - gpu-dev-network

    # Resource limits
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

  # AMD ROCm Development Environment
  rocm-dev:
    build:
      context: ../
      dockerfile: docker/rocm/Dockerfile
    image: gpu-programming-101:rocm
    container_name: gpu101-rocm-dev
    hostname: rocm-dev

    # GPU access for AMD
    devices:
      - "/dev/kfd:/dev/kfd"
      - "/dev/dri:/dev/dri"

    # Security options for ROCm
    security_opt:
      - seccomp:unconfined

    # Volume mounts
    volumes:
      - ../:/workspace/gpu-programming-101:rw
      - rocm-home:/root
      - rocm-cache:/root/.cache

    # Interactive mode
    stdin_open: true
    tty: true

    # Working directory
    working_dir: /workspace/gpu-programming-101

    # Network
    networks:
      - gpu-dev-network

    # Environment variables
    environment:
      - HIP_VISIBLE_DEVICES=0
      - HSA_OVERRIDE_GFX_VERSION=11.0.0
      - ROCM_VERSION=7.0

  # Development tools container (CPU-only for general development)
  dev-tools:
    build:
      context: ../
      dockerfile: docker/cuda/Dockerfile
      target: dev-base
    image: gpu-programming-101:dev-tools
    container_name: gpu101-dev-tools
    hostname: dev-tools

    # No GPU access - for documentation, build scripts, etc.

    # Volume mounts
    volumes:
      - ../:/workspace/gpu-programming-101:rw
      - dev-home:/root

    # Interactive mode
    stdin_open: true
    tty: true

    # Working directory
    working_dir: /workspace/gpu-programming-101

    # Network
    networks:
      - gpu-dev-network

# Named volumes for persistent data
volumes:
  cuda-home:
    driver: local
  cuda-cache:
    driver: local
  rocm-home:
    driver: local
  rocm-cache:
    driver: local
  dev-home:
    driver: local

# Custom network
networks:
  gpu-dev-network:
    driver: bridge
    ipam:
      config:
        - subnet: 172.20.0.0/16