Performance Reporting of Mathematical Library Installations with LAAB - An Overview
2026-08-13 • Performance
PerformanceMathematical Software
AI summaryⓘ
The authors created the LAAB framework to measure how well math libraries perform on high-performance computers, which are important for scientific work. They focus on making sure performance reports are clear about which software versions were used, how results relate to real scientific problems, how to handle measurement differences, and making reports easy to share and check. Their framework helps provide consistent and useful performance information for scientific computing users. This can help predict how much computer time is needed and improve resource planning.
mathematical librarieshigh-performance computingperformance benchmarksexecution settingsmeasurement variabilityresource allocationscientific applicationstraceabilitybenchmarking framework
Authors
Aravind Sankaran, Paolo Bientinesi
Abstract
We present the Linear Algebra Aware Benchmarks (LAAB) framework for systematically assessing and reporting the performance of mathematical library installations on HPC systems. Mathematical libraries provide interfaces for operations that form the computational building blocks of scientific applications. Reporting their performance is important for assessing application efficiency, estimating compute-time requirements, and preparing resource-allocation requests. In this paper, we define four objectives for performance reporting: 1) traceability, linking each report to the exact library installation and execution settings; 2) compatibility, relating library-operation performance to higher-level scientific applications that use them; 3) reliability, supporting interpretation in the presence of measurement variability; and 4) accessibility, ensuring that reports, benchmark definitions, and relevant metadata are available for inspection and reproduction. We then present the design of LAAB and show how it addresses the challenges associated with these objectives.