Skip to content

feat: add GPU support to metaflow-dev minikube setup #2609

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
Open

feat: add GPU support to metaflow-dev minikube setup #2609

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
dc75b62
feat: add GPU support to metaflow-dev minikube setup
cnaples79 Sep 17, 2025
20ec83b
Address Docker compatibility feedback for GPU support
cnaples79 Sep 17, 2025
8ca6ae4
minikube GPU: remove unsupported --devices flag; always use --gpus al…
cnaples79 Sep 17, 2025
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
32 changes: 32 additions & 0 deletions devtools/Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,9 @@ help:
@echo " down - Stop and clean up the environment"
@echo " all-up - Start the development environment with all services"
@echo " help - Show this help message"
@echo ""
@echo "Environment variables:"
@echo " MINIKUBE_ENABLE_GPU=auto|true|false - Control GPU support (default: auto)"

HELM_VERSION := v3.14.0
MINIKUBE_VERSION := v1.32.0
Expand All @@ -31,6 +34,7 @@ MAKE_CMD := $(MAKE) -f "$(MKFILE_PATH)"
MINIKUBE_CPUS ?= 4
MINIKUBE_MEMORY ?= 6144
MINIKUBE_DISK_SIZE ?= 20g
MINIKUBE_ENABLE_GPU ?= auto

ifeq ($(shell uname), Darwin)
minikube_os = darwin
Expand All @@ -48,6 +52,32 @@ else
tilt_arch = arm64
endif

# GPU detection for minikube
ifeq ($(MINIKUBE_ENABLE_GPU), auto)
# Auto-detect GPU availability
ifeq ($(shell command -v nvidia-smi >/dev/null 2>&1 && echo "nvidia"), nvidia)
gpu_detected = true
else ifeq ($(shell command -v rocm-smi >/dev/null 2>&1 && echo "amd"), amd)
gpu_detected = true
else
gpu_detected = false
endif
else ifeq ($(MINIKUBE_ENABLE_GPU), true)
# Force enable GPU support
gpu_detected = true
else
# Explicitly disabled
gpu_detected = false
endif

# Determine GPU flags for minikube
# Note: minikube supports `--gpus all` but does not accept `--devices ...`.
ifeq ($(gpu_detected), true)
gpu_flags = --gpus all
else
gpu_flags =
endif

# TODO: Move scripts to a folder

install-helm:
Expand Down Expand Up @@ -136,11 +166,13 @@ setup-minikube:
@echo "🔧 Setting up Minikube $(MINIKUBE_VERSION) cluster..."
@if ! $(MINIKUBE) status >/dev/null 2>&1; then \
echo "🚀 Starting new Minikube $(MINIKUBE_VERSION) cluster..."; \
if [ -n "$(gpu_flags)" ]; then echo "🎮 GPU support detected - enabling GPU passthrough"; else echo "💻 No GPU detected - starting without GPU support"; fi; \
$(MINIKUBE) start \
--cpus $(MINIKUBE_CPUS) \
--memory $(MINIKUBE_MEMORY) \
--disk-size $(MINIKUBE_DISK_SIZE) \
--driver docker \
$(gpu_flags) \
|| { echo "❌ Failed to start Minikube (check if Docker is running)"; exit 1; }; \
echo "🔌 Enabling metrics-server and dashboard (quietly)..."; \
$(MINIKUBE) addons enable metrics-server >/dev/null 2>&1; \
Expand Down