Open source tools for GPUs.

Three tools. Learn locally, monitor your cluster, benchmark your hardware. All free, all open source.

Free & Open Source

DeepLM Dev

The easiest way to build, learn, and interact with local AI. Run a model on the machine you already own, ask it questions about your hardware, and learn what fits before you download anything.

Free & Open Source

DeepLM Insights

Real-time Grafana dashboards and Prometheus metrics for HPC/SLURM GPU clusters. Track job performance, GPU utilization, power consumption, and checkpoint efficiency across your entire fleet — setup takes minutes.

5 Grafana dashboards included

Job Insights
Per-job CPU/GPU utilization, memory, priority, and power consumption
System Overview
Cluster-wide power, GPU temperature, fan speed, CPU/memory per node
Live Jobs
Real-time active job monitoring with 5-second refresh
Historical Jobs
Job duration analysis, completion rates, CPU hours by user
Checkpoint Analysis
Sync vs async checkpoint strategy comparison — stall time and overhead

How it works

  • SLURM prologue/epilogue hooks collect per-job metrics at start and end
  • Flask metrics API exposes Prometheus-format data on :5000/metrics
  • Cassandra stores job history for long-term trend analysis
  • Optional NVIDIA BCM integration for real GPU power data (falls back to TDP estimate)
  • Full stack deployable in one command via Docker Compose

Stack

SLURM hooks→ Flask metrics API→ Cassandra→ Prometheus→ Grafana
Coming Soon

DeepLM Optimize

Automatic cross-vendor GPU scheduling and workload optimization. See where your cluster is wasting capacity and let DeepLM fix it.

Free & Open Source

DeepLM Baseline

Multi-stage performance benchmarking for GPU clusters. Run three targeted test suites — compute, interconnect, and network — to establish ground-truth baselines before and after any infrastructure change. Structured JSON output for every stage.

Know exactly where your cluster stands against vendor specs. Find the GPUs that aren't pulling their weight, the NVLink bridges that are misconfigured, and the nodes that silently kill distributed training performance.