TY - GEN
T1 - Quantifying the Semantic Gap between Serial and Parallel Programming
AU - Zhang, Xiaochun
AU - Jones, Timothy M.
AU - Campanoni, Simone
N1 - Funding Information:
This work was supported by the Engineering and Physical Sciences Research Council (EPSRC), through grant references EP/K026399/1 and EP/P020011/1. This work was also made possible by support from the National Science Foundation via the awards CCF-1908488, CCF-2107042, CCF-2118708, CNS-1763743, PPoSS-2119069, and PPoSS-2028851. Additional data related to this publication is available in the data repository at https://doi.org/10.17863/CAM.76224.
Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Automatic parallelizing compilers are often constrained in their transformations because they must conservatively respect data dependences within the program. Developers, on the other hand, often take advantage of domain-specific knowledge to apply transformations that modify data dependences but respect the application's semantics. This creates a semantic gap between the parallelism extracted automatically by compilers and manually by developers. Although prior work has proposed programming language extensions to close this semantic gap, their relative contribution is unclear and it is uncertain whether compilers can actually achieve the same performance as manually parallelized code when using them. We quantify this semantic gap in a set of sequential and parallel programs and leverage these existing programming-language extensions to empirically measure the impact of closing it for an automatic parallelizing compiler. This lets us achieve an average speedup of 12.6× on an Intel-based 28-core machine, matching the speedup obtained by the manually parallelized code. Further, we apply these extensions to widely used sequential system tools, obtaining 7.1× speedup on the same system.
AB - Automatic parallelizing compilers are often constrained in their transformations because they must conservatively respect data dependences within the program. Developers, on the other hand, often take advantage of domain-specific knowledge to apply transformations that modify data dependences but respect the application's semantics. This creates a semantic gap between the parallelism extracted automatically by compilers and manually by developers. Although prior work has proposed programming language extensions to close this semantic gap, their relative contribution is unclear and it is uncertain whether compilers can actually achieve the same performance as manually parallelized code when using them. We quantify this semantic gap in a set of sequential and parallel programs and leverage these existing programming-language extensions to empirically measure the impact of closing it for an automatic parallelizing compiler. This lets us achieve an average speedup of 12.6× on an Intel-based 28-core machine, matching the speedup obtained by the manually parallelized code. Further, we apply these extensions to widely used sequential system tools, obtaining 7.1× speedup on the same system.
UR - http://www.scopus.com/inward/record.url?scp=85125195343&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85125195343&partnerID=8YFLogxK
U2 - 10.1109/IISWC53511.2021.00024
DO - 10.1109/IISWC53511.2021.00024
M3 - Conference contribution
AN - SCOPUS:85125195343
T3 - Proceedings - 2021 IEEE International Symposium on Workload Characterization, IISWC 2021
SP - 151
EP - 162
BT - Proceedings - 2021 IEEE International Symposium on Workload Characterization, IISWC 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 17th IEEE International Symposium on Workload Characterization, IISWC 2021
Y2 - 7 November 2021 through 9 November 2021
ER -