Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Porting numerical integration codes from CUDA to oneAPI: a case study

Conference · · No journal information
OSTI ID:1969670

We present our experience in porting optimized CUDA implementations to oneAPI. We focus on the use case of numerical integration, particularly the CUDA implementations of PAGANI and $$m$$-Cubes. We faced several challenges that caused performance degradation in the oneAPI ports. These include differences in utilized registers per thread, compiler optimizations, and mappings of CUDA library calls to oneAPI equivalents. After addressing those challenges, we tested both the PAGANI and m-Cubes integrators on numerous integrands of various characteristics. To evaluate the quality of the ports, we collected performance metrics of the CUDA and oneAPI implementations on the Nvidia V100 GPU. We found that the oneAPI ports often achieve comparable performance to the CUDA versions, and that they are at most 10% slower.

Research Organization:
Old Dominion U. (main); NVIDIA, Santa Clara; Fermi National Accelerator Laboratory (FNAL), Batavia, IL (United States)
Sponsoring Organization:
US Department of Energy
DOE Contract Number:
AC02-07CH11359
OSTI ID:
1969670
Report Number(s):
FERMILAB-CONF-23-007-LDRD-SCD; oai:inspirehep.net:2643005; arXiv:2302.05730
Journal Information:
No journal information, Journal Name: No journal information
Country of Publication:
United States
Language:
English

Similar Records

m-CUBES An efficient and portable implementation of multi-dimensional integration for gpus
Conference · Wed Feb 02 23:00:00 EST 2022 · OSTI ID:1844787

A case study of CUDA FORTRAN and OpenACC for an atmospheric climate kernel
Journal Article · Sat Apr 18 00:00:00 EDT 2015 · Journal of Computational Science · OSTI ID:1462913

Experience of Migrating a Parallel Graph Coloring Program from CUDA to SYCL
Technical Report · Fri Apr 01 00:00:00 EDT 2022 · OSTI ID:1864412

Related Subjects