[k8s] L40 GPUs get detected as L4s

[romilbhardwaj]

User reported L40 GPUs in their cluster were being shown as L4 GPUs in sky show-gpus

sky show-gpus:

Kubernetes GPUs (context: default)
GPU  REQUESTABLE_QTY_PER_NODE  TOTAL_GPUS  TOTAL_FREE_GPUS
L4   1, 2, 4                   4           4
Kubernetes per node GPU availability
NODE_NAME     GPU_NAME  TOTAL_GPUS  FREE_GPUS
node          L4        4           4

Labels:

                    nvidia.com/cuda.driver-version.major=535
                    nvidia.com/cuda.driver-version.minor=104
                    nvidia.com/cuda.driver-version.revision=12
                    nvidia.com/cuda.driver.major=535
                    nvidia.com/cuda.driver.minor=104
                    nvidia.com/cuda.driver.rev=12
                    nvidia.com/cuda.runtime-version.full=12.2
                    nvidia.com/cuda.runtime-version.major=12
                    nvidia.com/cuda.runtime-version.minor=2
                    nvidia.com/cuda.runtime.major=12
                    nvidia.com/cuda.runtime.minor=2
                    nvidia.com/gfd.timestamp=1732465495
                    nvidia.com/gpu-driver-upgrade-state=upgrade-done
                    nvidia.com/gpu.compute.major=8
                    nvidia.com/gpu.compute.minor=9
                    nvidia.com/gpu.count=4
                    nvidia.com/gpu.deploy.container-toolkit=true
                    nvidia.com/gpu.deploy.dcgm=true
                    nvidia.com/gpu.deploy.dcgm-exporter=true
                    nvidia.com/gpu.deploy.device-plugin=true
                    nvidia.com/gpu.deploy.driver=pre-installed
                    nvidia.com/gpu.deploy.gpu-feature-discovery=true
                    nvidia.com/gpu.deploy.node-status-exporter=true
                    nvidia.com/gpu.deploy.operator-validator=true
                    nvidia.com/gpu.family=ampere
                    nvidia.com/gpu.memory=46068
                    nvidia.com/gpu.mode=compute
                    nvidia.com/gpu.present=true
                    nvidia.com/gpu.product=NVIDIA-L40
                    nvidia.com/gpu.replicas=1
                    nvidia.com/gpu.sharing-strategy=none
                    nvidia.com/mig.capable=false
                    nvidia.com/mig.strategy=single
                    nvidia.com/mps.capable=false
                    nvidia.com/vgpu.present=false

The canonical_name in value here is likely the culprit, since the substring gets matched:

skypilot/sky/provision/kubernetes/utils.py

Lines 338 to 352 in ef2233b

	def get_accelerator_from_label_value(cls, value: str) -> str:
	"""Searches against a canonical list of NVIDIA GPUs and pattern
	matches the canonical GPU name against the GFD label.
	"""
	canonical_gpu_names = [
	'A100-80GB', 'A100', 'A10G', 'H100', 'K80', 'M60', 'T4g', 'T4',
	'V100', 'A10', 'P4000', 'P100', 'P40', 'P4', 'L4'
	]
	for canonical_name in canonical_gpu_names:
	# A100-80G accelerator is A100-SXM-80GB or A100-PCIE-80GB
	if canonical_name == 'A100-80GB' and re.search(
	r'A100.*-80GB', value):
	return canonical_name
	elif canonical_name in value:
	return canonical_name