NVIDIA DCGM

GPU management and monitoring software for NVIDIA data-center GPUs on Linux, with Kubernetes telemetry and an Apache 2.0 open-source core.

NVIDIA DCGM monitors and manages NVIDIA data-center GPUs on your own Linux servers. It's for infrastructure teams running GPU clusters, including those hosting AI workloads, who need to track hardware health, investigate slow jobs and control power use. It supports x86_64 and aarch64 (SBSA) systems.

DCGM collects GPU health, performance and utilization data to help teams distinguish hardware faults from application performance problems. Its low-overhead health checks are designed to run alongside active jobs. Active and passive diagnostics help identify failures, performance degradation and power inefficiencies, while system alerts flag problems that need attention.

Management includes power and clock policies, so teams can govern GPU behavior as well as observe it. DCGM works on its own or alongside cluster management and resource scheduling software. DCGM-Exporter exposes GPU metrics and health data in Kubernetes environments for monitoring and alerting; supported integrations include Prometheus, Bright Cluster Manager and IBM Spectrum LSF.

DCGM has an open-core architecture. Its foundational libraries and building blocks are open source under Apache 2.0, while some diagnostics and tests remain proprietary. Teams building their own management software can use its C and Python APIs or Go bindings.

GPU Management and Monitoring Kubernetes and Cloud-Native AI Linux

Are you the owner or maker of NVIDIA DCGM?

Claim this page with an email at developer.nvidia.com. NVIDIA DCGM gets the verified badge, and you can upgrade the listing to be featured on localhosted. Proud to be listed? Put our badge on your site.

Want more people to find NVIDIA DCGM?Promote it

Something wrong or outdated on this page?

Similar to NVIDIA DCGM

View all tools

Similar to NVIDIA DCGM

dstack

Self-hosted AI compute orchestration under MPL-2.0 for training and inference on GPU clouds, Kubernetes, VMs and bare-metal servers.

2.3KUpdated 1 day agoMPL-2.0

macOS · Windows · Linux · Docker#Agent Skills#Batch processing#Multi-user access

dstack is a self-hosted orchestration tool for AI teams managing compute across GPU clouds and their own servers. It puts cluster management, training jobs and model inference behind one interface, so teams can use different providers and accelerators without maintaining a separate workflow for each environment. It's open source under the Mozilla Public License 2.0.

Fine-Tuning and TrainingGPU Management and MonitoringInference Servers and ClustersKubernetes and Cloud-Native AI

LLMKube

Self-hosted LLM inference for Kubernetes with NVIDIA, AMD and Apple Silicon support, OpenAI-compatible APIs, and an Apache 2.0 license.

223Updated 16 hours agoApache-2.0

macOS · Linux#GGUF#Git integration#Guardrails

ROCm

An open-source GPU computing platform for AI training and inference on AMD hardware, with Linux and Windows support and PyTorch, JAX, vLLM and SGLang.

6.8KUpdated 4 weeks agoMIT

Windows · Linux

ROCm is AMD's open-source GPU computing platform for developers running AI training, inference and scientific workloads on their own hardware or servers. It supports selected Linux and Windows configurations on AMD Instinct, Radeon and Ryzen AI devices. Check the version-specific GPU, operating-system, driver and firmware compatibility matrix before installing. It's the software foundation for applications that need AMD GPU acceleration, including local LLM workloads.

AIBrix

Open-source LLM serving infrastructure for Kubernetes with multi-node inference, demand-based autoscaling, LoRA management and vLLM integration.

5.1KUpdated 20 hours agoApache-2.0

#Batch processing#Distributed execution#LoRA

ClearML

Self-hostable MLOps software for experiment tracking, pipelines, dataset versioning and model serving, with an Apache 2.0 Python SDK.

6.9KUpdated 2 days agoApache-2.0

Docker · Web#Code execution#Git integration#Multi-user access

Kaito

A self-hosted Kubernetes operator that deploys Hugging Face models with vLLM, manages GPU capacity, and runs fine-tuning and document retrieval services.

1KUpdated 7 days ago

#Distributed execution#Hugging Face integration#LoRA

Kaito manages self-hosted LLM inference, fine-tuning, and document retrieval services in a Kubernetes cluster. It's for teams that want to run models on infrastructure they control while reducing the work of sizing GPU resources and managing model deployments. The project is open source under Apache 2.0.