Installation
This guide covers everything you need to install and set up ncprof, including prerequisites, the VS Code extension, and the ncompass SDK.
Prerequisites
Before installing ncprof, ensure you have the following:
Docker & Docker Compose
The profiling backend runs as Docker containers, so Docker is required. Check if Docker is installed:
docker --version
docker compose versionInstallation:
- Windows/macOS: Docker Desktop
- Linux: Official docs
VS Code 1.80.0+
Check your VS Code version:
- Open VS Code
- Go to Help → About
- Ensure version is 1.80.0 or higher
If needed, download the latest version from code.visualstudio.com .
Installing the VS Code Extension
Install from VSCode Marketplace
The easiest way to install the extension is directly from the VSCode Marketplace:
- Open VS Code
- Go to the Extensions view (
Ctrl+Shift+XorCmd+Shift+X) - Search for “nCompass” or “ncprof” in the Extensions search bar
- Click Install on the “nCompass Performance Optimization IDE” extension
Alternatively, you can install directly from the VSCode Marketplace .
Initial Setup Prompts
On startup, the extension will prompt for 2 details:
- User Email - We don’t collect any trace or code data, but we do record tool crash logs along side the user email to aid us with fixing bugs.
- Trace Directory - This is a path on your local system which you are currently storing traces or plan on storing traces to. This is so that the tool knows where to look for traces to analyze.
Verify Extension Installation
- Open VS Code
- Go to the Extensions view (
Ctrl+Shift+XorCmd+Shift+X) - Search for “ncprof” or “NCProf GPU Trace Viewer”
- You should see the extension listed as installed
Installation Video Guide
Watch our Installation Guide for a step-by-step walkthrough.
Note: Please ignore any mentions of installing from VSIX in this video. You can install directly from the VSCode Marketplace instead.
For example code showing how to generate trace files, check out our examples repository .
Installing the ncompass SDK
The ncompass SDK is the Python library that instruments your code based on ncprof’s configuration.
Install via pip
pip install ncompassVerify Installation
python -c "import ncompass; print(f'ncompass version: {ncompass.__version__}')"You should see the installed version number without any errors.
Initial Setup
First-Time Backend Launch
The first time you use ncprof, you need to start the profiling backend:
- Open VS Code in your project workspace
- Press
Ctrl+Shift+P(Windows/Linux) orCmd+Shift+P(Mac) - Type “NCProf: Start Profiler Backend (Docker)”
- Press Enter
The backend will:
- Create a
.cache/ncompassdirectory in your workspace - Pull the necessary Docker images (this may take a few minutes the first time)
- Start Redis and the profiling API server containers
You should see a success notification once the backend is running.
Configure Auto-Start (Optional)
To automatically start the backend when you open VS Code:
- Open VS Code Settings (
Ctrl+,orCmd+,) - Search for “ncprof”
- Enable
ncprof.backend.autoStart
Or add to your settings.json:
{
"ncprof.backend.autoStart": true
}Verify Backend Status
Check that the backend is running correctly:
- Press
Ctrl+Shift+P(Windows/Linux) orCmd+Shift+P(Mac) - Type “NCProf: Check NCProf Backend Status”
- Press Enter
You should see output in the terminal showing running containers.
Updating
Update the Extension
The extension will automatically update through VS Code’s extension marketplace. To manually check for updates:
- Open Extensions view (
Ctrl+Shift+XorCmd+Shift+X) - Search for “nCompass” or “NCProf GPU Trace Viewer”
- If an update is available, you’ll see an “Update” button
- Click “Update” to install the latest version
Alternatively, you can visit the VSCode Marketplace to see the latest version.
Update the ncompass SDK
pip install --upgrade ncompassUninstalling
Remove the Extension
- Open Extensions view (
Ctrl+Shift+X) - Find “NCProf GPU Trace Viewer”
- Click Uninstall
Remove the ncompass SDK
pip uninstall ncompassRemove Workspace Data
# Remove the cache directory
rm -rf .cache/ncompassNext Steps
Now that you have ncprof installed:
- Quick Start: Complete the 5-minute getting started guide
- Detailed breakdown of profiling workflow: Learn the detailed profiling workflow
- Settings & Commands: Explore available configuration options