BibTeX Export
@INPROCEEDINGS{, author = {Schl{\"{u}}tter, Marc and Philippen, Peter and Morin, Laurent and Geimer, Markus and Mohr, Bernd}, editor = {Bader, Michael and Bode, Arndt and Bungartz, Hans-Joachim and Gerndt, Michael and Joubert, Gerhard R. and Peters, Frans J.}, keywords = {accelerator, CUBE, GPGPU, GPU, OpenACC, optimisation, performance, PHMPP, profiling, Score-P, tools, tracing}, month = mar, title = {Profiling Hybrid HMPP Applications with Score-P on Heterogeneous Hardware}, booktitle = {Parallel Computing: Accelerating Computational Science and Engineering (CSE)}, series = {Advances in Parallel Computing}, volume = {25}, year = {2014}, pages = {773 - 782}, publisher = {IOS Press}, isbn = {978-1-61499-381-0}, url = {http://www.ebooks.iospress.nl/volumearticle/35952}, doi = {10.3233/978-1-61499-381-0-773}, abstract = {In heterogeneous environments with multi-core systems and accelerators, programming and optimizing large parallel applications turns into a time-intensive and hardware-dependent challenge. To assist application developers in this process, a number of tools and high-level compilers have been developed. Directive-based programming models such as HMPP and OpenACC provide abstractions over low-level GPU programming models, such as CUDA or OpenCL. The compilers developed by CAPS automatically transform the pragma-annotated application code into low-level code, thereby allowing the parallelization and optimization for a given accelerator hardware. To analyze the performance of parallel applications, multiple partners in Germany and the US jointly develop the community measurement infrastructure Score-P. Score-P gathers performance execution profiles, which can be presented and analyzed within the CUBE result browser, and collects detailed event traces to be processed by post-mortem analysis tools such as Scalasca and Vampir. In this paper we present the integration and combined use of Score-P and the CAPS compilers as one approach to efficiently parallelize and optimize codes. Specifically, we describe the PHMPP profiling interface, it's implementation in Score-P, and the presentation of preliminary results in CUBE.} }
Copy