skills.rest
GindaChen
GindaChenCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

nsys-ai

Analyze NVIDIA Nsight Systems profiles to identify GPU bottlenecks and propose fixes.

71|20|Updated Feb 26, 2026
Search Tags:#diff#root-cause#profile-analysis#nsys-ai#gpu-bottlenecks#nvidia-nsys
One-click install
npx skills add https://github.com/GindaChen/nsys-ai --skill nsys-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: nsys-ai
Source: https://github.com/GindaChen/nsys-ai/tree/main/skills/analyze
Command: npx skills add https://github.com/GindaChen/nsys-ai --skill nsys-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

nsys-ai analyzes NVIDIA Nsight Systems profiles (.sqlite or .nsys-rep) to identify GPU bottlenecks and deliver actionable fixes, enabling faster debugging and performance improvement.

Core Features & Use Cases

  • Root-cause diagnosis: identifies bottlenecks such as NCCL contention, idle/gap issues, and MFU inefficiencies; suggests concrete fixes.
  • Diff & comparison: compares two profiles to explain regressions or improvements.
  • Guided delivery: follows universal principles to maintain evidence-based conclusions and code-change formatting in fixes.
  • Use Case: profile-driven optimization for multi-GPU workloads in DL training or GPU-accelerated workloads.

Quick Start

Provide a Nsight Systems profile (.sqlite or .nsys-rep) and run the skill to receive a root-cause and actionable fix.

Frequently Asked Questions about nsys-ai

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

FAQPage Schema
How do I identify GPU bottlenecks from an NVIDIA Nsight Systems profile?▼

To identify GPU bottlenecks from an NVIDIA Nsight Systems profile, provide the .sqlite or .nsys-rep file for analysis to receive a concrete root-cause diagnosis and actionable fix.

What is the best way to diagnose NCCL contention and idle gaps in multi-GPU workloads?▼

Diagnosing NCCL contention and idle gaps in multi-GPU workloads requires analyzing Nsight Systems profiles to pinpoint MFU inefficiencies and variance, yielding specific code-level fixes.

Can I compare two Nsight Systems profiles to explain a performance regression?▼

Yes, you can diff two Nsight Systems profiles to explain performance regressions or improvements by comparing their .sqlite or .nsys-rep data to isolate the root-cause changes.

Does the bottleneck analysis provide code examples for fixing GPU inefficiencies?▼

Yes, the bottleneck analysis provides code examples for fixing GPU inefficiencies by applying official mode references and principles to deliver evidence-based, actionable code changes.

When do I need to analyze a .nsys-rep file for deep learning training optimization?▼

You need to analyze a .nsys-rep file for deep learning training optimization when facing unresolved MFU drops or communication overhead and require profile-driven, actionable fixes to improve workload performance.

Related Skills

Explore More Data & Analytics Skills→
yo-stevenyo-steven

perf-nsight-systems

Profile CPU, GPU, and communication timelines from .nsys-rep captures.

Community
Intermediate
kindoflukekindofluke

profile-analyzer

Analyze Sirius GPU profiles from Nsight Systems to identify bottlenecks and regression risks.

Community
Advanced
LancerLabLancerLab

perf-nsight-compute-analysis

Analyze NVIDIA Nsight Compute reports to classify GPU kernel bottlenecks.

Official
Intermediate
xenitV1xenitV1
132

performance-profiling

Automate Lighthouse-based performance profiling to identify web application bottlenecks.

Community
Advanced