Utilyze: Open-Source GPU Monitor That Measures Real Compute Throughput, Not Just Kernel Activity
The standard GPU utilization metric used by nvidia-smi, nvtop, Weights & Biases, Amazon CloudWatch, Google Cloud Monitoring, and Azure Monitor is misleading. It reports the fraction of time any kernel is running, so a GPU can show 100% utilization while only using 1-10% of real compute capacity. Teams relying on this for capacity planning may think systems are saturated when they're actually underutilized.
Utilyze
SysTalize released Utilyze (utlz), an open-source (Apache 2.0) tool that measures GPU utilization differently. Instead of kernel activity, it samples hardware performance counters and reports compute and memory throughput relative to the hardware's theoretical limits. It also estimates an attainable utilization ceiling for a given workload.
Installation
curl -fsSL https://systalyze.com/utilyze/install.sh | bashUtilyze runs alongside any AI workload in real time with negligible overhead. In production deployments, it has revealed orders-of-magnitude performance headroom in systems that standard tools declared fully saturated.
Why This Matters
AI compute is scarce: H100 one-year rental contracts rose ~40% from October 2025 to March 2026, and lead times for GPUs stretch months. Wasted spend on unnecessary hardware and energy is massive. Accurate measurement is the prerequisite for optimization — every percentage point of real throughput recovered saves money and resources.
Check the GitHub repo: https://github.com/systalyze/utilyze
📖 Read the full source: HN LLM Tools
👀 See Also
Chrome Skills: Save and Reuse AI Prompts as One-Click Tools
Google's Chrome Skills feature lets users save AI prompts as reusable workflows that run with a single click on any webpage. Skills can be accessed by typing forward slash (/) or clicking the plus sign (+) in Gemini in Chrome.
Myelin: MD Extractor and Evaluator for Claude Code Procedural Memory
Myelin is an MCP server that hooks into Claude Code via PostToolUse to capture tool calls, automatically extracting .md procedure files from successful sessions and tracking whether Claude follows existing procedures step-by-step.
/goal for Claude Code: persistent tasks with adversarial review
A /goal command for Claude Code that keeps it working on a long task across many turns, with an optional separate Claude session reviewing the final result to prevent false completion.
Canopy: Terminal Dashboard for Managing Multiple Claude Code Agents
Canopy is an open source terminal UI that provides a single dashboard view for tracking multiple AI coding agents running across git worktrees. It shows agent states (running, idle, waiting for input, done, errored) and lets you jump into sessions or send input without fully switching.