Home Publications edited volumes Awards Research Teaching Miscellaneous Full CV [pdf] BLOG
Events
Past Events
|
Publications of Torsten Hoefler
H. Schweizer, Maciej Besta, Torsten Hoefler:
| | Evaluating the Cost of Atomic Operations on Modern Architectures
(presented in San Francisco, CA, USA, ACM, Oct. 2015, Accepted at the 24th International Conference on Parallel Architectures and Compilation (PACT'15) )
AbstractAtomic operations (atomics) such as Compare-and-Swap (CAS) or
Fetch-and-Add (FAA) are ubiquitous in parallel programming. Yet,
performance of these operations has not been thoroughly analyzed,
hindering the development and the reasoning of high performance
codes. In this paper we establish an evaluation methodology, develop a performance model, and present a set of detailed benchmarks for latency and bandwidth of different atomics. We consider
various state-of-the-art x86 architectures: Intel Haswell, Xeon Phi,
Ivy Bridge, and AMD Bulldozer. The results unveil surprising performance relationships between the considered atomics and architectural properties such as the coherence state of the accessed cache
lines. One key finding is that all the tested atomics have comparable latency and bandwidth even if they are characterized by different consensus numbers. Another insight is that the hardware implementation of atomics prevents any instruction-level parallelism
even if there are no dependencies between the issued operations. Finally, we discuss solutions to the discovered performance issues in
the analyzed architectures. Our analysis enables simpler and more
effective parallel programming and accelerates data processing on
various architectures deployed in both off-the-shelf machines and
large compute systems.
Documentsdownload article: download slides: | | BibTeX | @inproceedings{atomics-cost, author={H. Schweizer and Maciej Besta and Torsten Hoefler}, title={{Evaluating the Cost of Atomic Operations on Modern Architectures}}, year={2015}, month={Oct.}, location={San Francisco, CA, USA}, publisher={ACM}, note={Accepted at the 24th International Conference on Parallel Architectures and Compilation (PACT'15)}, source={http://www.unixer.de/~htor/publications/}, } |
|
|