system-profile

Profile scripts, services, or GPU workloads to identify performance bottlenecks.

1|1|Updated May 19, 2026

Search Tags:#profiling #instrumentation #gpu #performance bottleneck #memory analysis #nvidia-smi #nccl

Ready to run

npx skills add https://github.com/zhuyingqin/ARIS-WEB --skill system-profile-zhuyingqin

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: system-profile
Source: https://github.com/zhuyingqin/ARIS-WEB/tree/main/crates/runtime/assets/skills/system-profile
Command: npx skills add https://github.com/zhuyingqin/ARIS-WEB --skill system-profile-zhuyingqin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve?

It solves the problem of unclear performance bottlenecks by profiling a chosen target and producing structured results that point to actionable improvements.

Core Features & Use Cases

Targeted profiling for scripts, processes, and GPU systems: Profiles a script, running PID/service, or an end-to-end framework/system such as vLLM serving to localize where time, memory, and compute are spent.
Instrumentation-driven analysis when needed: Chooses external profiling tools and, when necessary, writes instrumentation to measure wall time vs CPU time, CPU↔GPU transfers, GPU kernel timing, and communication overhead.
Structured performance report with recommendations: Produces tables across CPU overhead, memory redundancy, interconnect/collectives, and GPU compute utilization, then ranks recommendations by expected impact.

Quick Start

Profile the target you care about (for example, a running process or a GPU service) and ask for a performance report with bottleneck causes and recommended fixes.

Frequently Asked Questions about system-profile

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I find GPU performance bottlenecks in a running service or script?▼

Profiling a running service or script identifies performance bottlenecks by measuring GPU compute efficiency, CPU↔GPU transfers, and NCCL communication overhead, outputting structured tables with ranked recommendations.

What is instrumentation-driven profiling and when do I need it for memory analysis?▼

Instrumentation-driven profiling inserts targeted code to measure wall time, CPU time, and memory redundancy when external tools are insufficient. You need it to pinpoint specific interconnect or compute inefficiencies within your workload.

How do I profile NCCL collectives and GPU-GPU communication overhead?▼

Profile NCCL collectives and GPU-GPU communication overhead by applying external profiling tools or writing custom instrumentation to measure interconnect behavior, collecting the outputs into a structured performance report.

Can I profile a vLLM serving framework to localize CPU overhead and memory usage?▼

Yes, you can profile an end-to-end framework like vLLM serving to localize where time, CPU overhead, and memory usage are spent. It generates structured tables ranking recommendations by expected impact.

What is the best way to analyze CPU overhead and memory redundancy without dedicated profiling tools?▼

The best way to analyze CPU overhead and memory redundancy without dedicated tools is writing targeted instrumentation to measure execution time and memory usage, then generating a structured changelog of modified files.

What limitations exist when profiling a system with custom instrumentation?▼

A limitation of profiling with custom instrumentation is the requirement to manually write and insert code, which produces a changelog of modified files. It depends on selecting suitable external tools when available to minimize overhead.