Getting started

Install PantheonGPU and run a first validation

PantheonGPU automatically detects CUDA, ROCm/HIP, or mock mode. Run the pantheon command directly after installation.

Install on Ubuntu or Debian

Install the basic build tools:

sudo apt-get update
sudo apt-get install -y make g++

Install the compiler for your platform. You only need one:

NVIDIA CUDAAMD ROCm/HIP

sudo apt-get install -y nvidia-cuda-toolkit

sudo apt-get install -y hipcc

Download and install the latest Debian package:

VERSION=1.0.16
wget "https://github.com/pantheongpu/pantheongpu_website/releases/download/v${VERSION}/pantheongpu_${VERSION}_amd64.deb"
sudo apt install "./pantheongpu_${VERSION}_amd64.deb"

Run a first test

Run a short inventory test:

pantheon --test baseline_metrics --duration 10

Then run a targeted stress test on GPU 0:

pantheon --test fp64_virus --duration 30 --gpu 0

Profiling and reports

Use --verify to validate workload output. Use --profile to collect performance counters, traces, and a per-workload HTML summary:

pantheon --test llm_decode --duration 60 --gpu 0 --verify --profile

See Test Documentation for workload guidance, Releases and Downloads for other installation methods, and Reports for output details.

Uninstall

Remove a Debian package installation with:

sudo apt-get remove pantheongpu

To also remove PantheonGPU runtime-created files and the current user's compiled workload cache, or to remove a portable installation on RHEL-family systems:

curl -fsSL https://pantheongpu.com/uninstall.sh | sudo sh

This leaves CUDA, ROCm, system compilers, and benchmark reports outside PantheonGPU's installation and cache directories untouched.