pytest-gpu-proof: Enabling Cloud-CPU Continuous Integration for GPU Code with Local GPU Attestation

2026-09-23

AI Summary

This study addresses the problem of insufficient code test coverage caused by the prohibitive costs of cloud-based GPU continuous integration (CI). To overcome this, we propose a trusted local execution framework implemented as a Pytest plugin. This approach enables GPU-dependent code to execute locally while employing digital signature mechanisms to verify result integrity, thereby facilitating seamless integration into low-cost CPU-based CI pipelines. The core innovation lies in leveraging cryptographic techniques to guarantee the trustworthiness of distributed execution, effectively balancing testing costs with coverage rates. The proposed tool has been open-sourced and deployed within our laboratory’s software stack, offering a lightweight, secure, and scalable continuous integration solution for AI systems development.

Abstract

GPU acceleration is now routine across robotics, but cloud-hosted GPU continuous integration (CI) runners are expensive, resulting in severe under-testing of GPU-accelerated code. We present pytest-gpu-proof, an open-source pytest plugin offering a practical middle ground. Tests can be run on a local machine, signed with a receipt of exactly what ran and what it produced, and integrated into standard CPU CI workflows (e.g., GitHub Actions). The tool is open source and on PyPI, and we are actively integrating it across our lab's software stack.

Problem

• GPU acceleration
• continuous integration
• cloud CI cost
• under-testing
• robotics

Innovation

• GPU attestation
• continuous integration
• pytest plugin
• cloud-CPU CI
• test signing