NVIDIA Cloud Native

docs, youtube, youtube developer

Graphical cards

Name	Specifications	Comments
T4
T400		Affordable
V100
A100

GPUs on Kubernetes

• Install Kubernetes
• Improving GPU Utilization in Kubernetes
• On Demand Webinar

NVIDIA Multi-Instance GPU (MIG)

Multi-Instance GPU (MIG) expands the performance and value of NVIDIA H100, A100, and A30 Tensor Core GPUs. MIG can partition the GPU into as many as seven instances, each fully isolated with its own high-bandwidth memory, cache, and compute cores.

→ product

NVIDIA GPU Cloud (NGC)

• NGC catalog

Components

NVIDIA Container Runtime

→ code

NVIDIA Container Toolkit

→ code

NVIDIA DCGM-Exporter

DCGM-Exporter exposes GPU metrics exporter for Prometheus leveraging NVIDIA Data Center GPU Manager (DCGM)

→ code, docs

NVIDIA GPU feature discovery

→ code

NVIDIA GPU Operator

→ code, docs

• Platform suppport
• Getting support

NVIDIA Device Plugin

→ code

Tutorials

NVIDIA GPU Operator in K3s

• Installing & Using the NVIDIA GPU Operator in K3s with Rancher by Virtual Thoughts - November 21, 2022

NVIDIA GPUs with SLES

ℹ Full official support should come in early 2023

• Build and push a driver image for SLES (from GitLab project)

git clone https://gitlab.com/nvidia/container-images/driver.git && cd sle15

docker build . -t path/to/your/repo/driver:515.65.01-sles15.3 \
  --build-arg DRIVER_VERSION=515.65.01 \
  --build-arg CUDA_VERSION=11.7.1 \
  --build-arg SLES_VERSION=15.3

docker push path/to/your/repo/driver:515.65.01-sles15.3

• Deploy GPU Operator and specify custom driver image

helm install gpu-operator nvidia/gpu-operator \
  --create-namespace -n gpu-operator \
  --set driver.repository=path/to/your/repo