LibraryX: A Framework for Cross-Library-Call Optimization

Rao, Sanil [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]; Prakash, Anant [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]; Zhang, Naifeng [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]; Mankad, Het [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]; Franchetti, Franz [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]

doi:10.1109/ipdps65963.2026.00097

DOE OSTI · 3377170

LibraryX: A Framework for Cross-Library-Call Optimization

Rao, Sanil [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]·Prakash, Anant [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]·Zhang, Naifeng [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]·Mankad, Het [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]·Franchetti, Franz [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]

View original source ↗Export metadata ↓

Abstract

Scientific applications utilize performance libraries as a software engineering concept: these libraries encapsulate important and well-understood (mathematical) operations, allow for reuse, and are implemented and tuned by experts. Domain scientists then implement complex algorithms based on these domainspecific libraries. While individual library calls are optimized, larger performance gains across sequences of calls—sometimes spanning multiple libraries—are often unrealized, forcing a trade-off between performance and implementation complexity.To overcome this issue, we propose LibraryX, an approach and a system that allows for cross-library-call optimization even when library calls stem from multiple performance libraries. LibraryX annotates library calls with semantic information and optimizes entire directed acyclic graphs (DAGs) of calls dynamically using the SPIRAL code generation system. We demonstrate its effectiveness across a range of memory bound workloads, achieving significant speedups on Nvidia, AMD, and Intel accelerators compared to code using native libraries without cross-call optimization.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Rao, Sanil [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA], Prakash, Anant [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA], Zhang, Naifeng [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA], Mankad, Het [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA], Franchetti, Franz [Carnegie Mellon University,Department of Electrical and Computer Engineering,Pittsburgh,PA,USA]. 2026-05-25. LibraryX: A Framework for Cross-Library-Call Optimization. https://doi.org/10.1109/ipdps65963.2026.00097

Cite the original work for its findings. Save a collection to share your selection of sources.