gpu-burn is a Python package that provides managed PaddlePaddle workloads to control and stress-test GPU compute utilization. It allows developers to burn GPU resources in a controlled manner for benchmarking, testing, or load simulation. Useful for machine learning engineers and infrastructure teams managing GPU clusters.
gpu-burn is a Developer Tools project. It focuses on controlling and measuring GPU utilization during compute-intensive workloads and testing. gpu-burn is an open-source project aimed at developers. gpu-burn is open source under the MIT license. It runs on the command line.
Behind gpu-burn is cangtianhuang, and it first shipped in 2026. Development happens publicly on GitHub with 1 commit in the last 90 days. Key capabilities include GPU Stress Testing, Utilization Control, and PaddlePaddle Workload.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do