Gpu memory safety evaluated with new standardized benchmark suite
Towards Standardized Evaluation of GPU Memory Safety with GMSBench
Cryptography and SecurityHardware Architecture
Summary
Memory bugs in programs that run on GPUs can cause crashes or security risks, but testing tools that find these bugs is hard without a common set of tests. The authors created GMSBench, a collection of 149 small CUDA tests that cover many types of memory errors in GPU programs. This benchmark helps see how well different tools catch these errors and reveals where they may miss problems. The authors used GMSBench to check one popular GPU debugging tool across several GPU types, showing how it can improve evaluation.
What this means in practice
- •For gpu software developers: Use the GMSBench suite to test and improve memory safety tools in GPU programs before deployment.
- •For hardware quality assurance teams: Assess GPU memory error detection capabilities across GPU architectures with a consistent test suite.
Authors
Saurabh Singh, Jaewon Lee, Seonjin Na, Hyesoon Kim
Abstract
As GPUs become increasingly integral to high-performance computing and machine learning, ensuring memory safety in GPU programs has become crucial for reliable and secure execution. However, evaluating GPU memory safety techniques remains challenging due to the lack of comprehensive and standardized benchmarks. In this paper, we present GMSBench, a GPU memory safety benchmark designed to evaluate a broad range of memory safety violations across different GPU memory spaces and execution scenarios. GMSBench comprises 149 self-contained CUDA tests spanning spatial, temporal, and concurrency errors. The suite provides a standardized foundation for the evaluation and comparative analysis of GPU memory safety mechanisms and helps expose gaps in their detection coverage. We demonstrate the utility of GMSBench by evaluating Compute Sanitizer, a widely used GPU memory error detection tool across multiple GPU architectures.