GPU Cluster Monitoring (GCM): Large-Scale AI Research Cluster Monitoring
-
Updated
Jul 9, 2026 - Python
#
dcgm
Here are 25 public repositories matching this topic...
GPU Observability with workload attribution. One OTLP agent per node ties hardware metrics (NVIDIA, AMD, Intel Gaudi) to the K8s pod or Slurm job burning the GPU.
kubernetes ai amd gpu helm slurm nvidia gpu-monitoring mlops opentelemetry llm-observability nvml-monitoring dcgm intel-gaudi-base-operator gpu-observability workload-attribution
-
Updated
Jul 7, 2026 - Python
Hands-on GPU/HPC infrastructure operations: K8s GPU scheduling, HAMi sharing, Slurm, observability & vLLM inference. Learn it free on a laptop; validate on one cheap GPU.
kubernetes devops hpc gpu grafana slurm inference prometheus nvidia learning-by-doing observability mlops k3s gpu-scheduling hami vllm ai-infrastructure gpu-sharing dcgm kai-scheduler
-
Updated
Jul 7, 2026 - Shell
Simulate NVIDIA GPUs for testing. 7 behavior profiles, scale to 1000+ GPUs, Docker-ready Prometheus exporter using DCGM
testing research monitoring simulation metrics gpu telemetry observability chaos-engineering gpu-monitoring dcgm-exporter dcgm
-
Updated
Jan 7, 2026 - Python
End-to-end observability for disaggregated LLM inference on EKS — DCGM metrics, KEDA autoscaling on GPU signals, per-namespace cost attribution, multi-agent OTel tracing, and Istio mTLS. Reference implementation for the OpenTelemetry AI Inference Platform blueprint.
-
Updated
Jul 2, 2026 - Shell
GPU-native agent-swarm orchestration for the NVIDIA AI stack — NeMo, NIM, Triton, DCGM, NGC, NIXL, OpenShell. Spawn GPU-pinned agent teams across DGX/HGX nodes with NVLink-aware scheduling, task DAGs, adaptive scheduling, and full observability.
python cli nim hpc gpu slurm orchestration nvidia hyperparameter-optimization triton nemo ai-agents mlops dgx nvlink llm agent-swarm dcgm
-
Updated
Mar 28, 2026 - Python
Automated acceptance toolkit for Linux deep learning GPU servers
-
Updated
Jun 18, 2026 - Python
Prometheus exporter for hardware telemetry from DMTF Redfish-capable BMCs. Multi-target probe pattern. Demo stack included.
go golang monitoring data-center ipmi sre observability redfish bmc hardware-monitoring dcgm prometheus-exporte
-
Updated
May 9, 2026 - Go
Open-source GPU dynamic power management for datacenter — Python brain, Rust agent, Prometheus/Grafana
-
Updated
Jun 30, 2026 - Python
Real-time TTFT, TPOT, ITL, E2EL benchmarking dashboard for vLLM — live animated charts, Prometheus scrape, NVIDIA DCGM GPU metrics, and inference bottleneck diagnostics.
python benchmarking real-time dashboard grafana prometheus observability nvidia-gpu mlops ttft llmops vllm llm-inference dcgm
-
Updated
Jun 30, 2026 - HTML
kubectl plugin that compares requested GPU resources against DCGM Exporter utilization metrics and generates rightsizing recommendations with projected monthly cost savings. Supports nvidia.com/gpu and amd.com/gpu — the gap VPA leaves open.
go kubernetes gpu nvidia cloud-native kubectl resource-management cost-optimization finops amd-gpu mlops kubectl-plugin rightsizing ai-infrastructure dcgm
-
Updated
Apr 23, 2026 - Go
GPU-aware scheduling, DCGM observability, and KEDA autoscaling for LLM inference (vLLM) on Kubernetes — runs on a laptop with simulated GPUs
-
Updated
Jun 12, 2026 - Makefile
Production-grade health monitoring and predictive fault management system for NVIDIA A100/H100 GPU fleets
machine-learning kafka monitoring gpu nvidia predictive-maintenance failure-prediction timescaledb health-monitoring a100 h100 dcgm
-
Updated
Feb 20, 2026 - Python
Kubernetes controller that scrapes NVIDIA DCGM exporter for per-node GPU health and auto-cordons nodes throwing XID errors or running too hot. Closed-loop reliability tooling for GPU clusters.
kubernetes golang gpu site-reliability-engineering prometheus nvidia sre kubernetes-operator kubebuilder ai-infrastructure operator-pattern dcgm
-
Updated
May 8, 2026 - Go
GPU workload analyzer for AI infra teams on Kubernetes — semantic problem detection beyond raw metrics
-
Updated
May 19, 2026 - Python
See the Model FLOPs Utilization gap behind your GPU's "100% utilization": the GPU spend nvidia-smi and DCGM hide. No root access required.
gpu cuda prometheus nvidia observability mfu gpu-monitoring finops gpu-utilization llm llm-serving vllm llm-inference sglang dcgm model-flops-utilization gpu-efficiency
-
Updated
Jun 23, 2026 - Go
nvidia dcgm exporter container only
-
Updated
Feb 9, 2026 - Go
Prometheus + Grafana + NVIDIA DCGM Exporter + Alertmanager stack for monitoring CPU, memory, and GPU utilization with email alerting, all via Docker Compose
devops monitoring docker-compose grafana prometheus nvidia alertmanager observability gpu-monitoring dcgm
-
Updated
Jul 10, 2026
Mock NVIDIA dcgm-exporter with MIG metrics simulation for development and testing
-
Updated
May 29, 2026 - Go
Improve this page
Add a description, image, and links to the dcgm topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the dcgm topic, visit your repo's landing page and select "manage topics."