{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T20:24:05Z","timestamp":1791318245508,"version":"4.3.1"},"reference-count":38,"publisher":"SAGE Publications","issue":"2","license":[{"start":{"date-parts":[[2013,10,11]],"date-time":"2013-10-11T00:00:00Z","timestamp":1381449600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["The International Journal of High Performance Computing Applications"],"published-print":{"date-parts":[[2014,5]]},"abstract":"<jats:p>Graphics processing units (GPUs) have become an important platform for general-purpose computing, thanks to their high parallel throughput and high memory bandwidth. GPUs present significantly different architectures from CPUs and require specific mappings and optimizations to achieve high performance. This makes GPU workloads demonstrate application characteristics different from those of CPU workloads. It is critical for researchers to understand the first-order metrics that most influence GPU performance and scalability. Furthermore, methodologies and associated tools are needed to analyze and predict the performance of GPU applications and help guide users\u2019 purchasing decisions.<\/jats:p>\n                  <jats:p>In this work, we study the approach of predicting the performance of GPU applications by correlating them to existing workloads. One tenet of benchmark design, also a motivation of this paper, is that users should be given the capability to leverage standard workloads to infer the performance of applications of their interest. We first identify a set of important GPU application characteristics and then use them to predict performance of an arbitrary application by determining its most similar proxy benchmarks. We demonstrate the prediction methodology and conduct predictions with benchmarks from different suites to achieve better workload coverage. The experimental results show that we are able to achieve satisfactory performance predictions, although errors are higher for outlier applications. Finally, we discuss several considerations for systematically constructing future benchmark suites.<\/jats:p>","DOI":"10.1177\/1094342013507960","type":"journal-article","created":{"date-parts":[[2013,10,11]],"date-time":"2013-10-11T23:27:43Z","timestamp":1381534063000},"page":"238-250","update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":21,"title":["BenchFriend"],"prefix":"10.1177","volume":"28","author":[{"given":"Shuai","family":"Che","sequence":"first","affiliation":[{"name":"AMD Research, CA, USA"},{"name":"Most of this work was completed while Shuai Che was with the University of Virginia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kevin","family":"Skadron","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Virginia, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2013,10,11]]},"reference":[{"key":"bibr1-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1693453.1693470"},{"key":"bibr2-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/ISPASS.2009.4919648"},{"key":"bibr3-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1454115.1454128"},{"key":"bibr4-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/SC.2005.33"},{"key":"bibr5-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2009.5306797"},{"key":"bibr6-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2010.5650274"},{"key":"bibr7-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1735688.1735702"},{"key":"bibr8-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2010.5649549"},{"key":"bibr9-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/HICSS.1998.649247"},{"key":"bibr10-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2011.6114195"},{"key":"bibr11-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1555754.1555775"},{"key":"bibr12-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2007.56"},{"key":"bibr13-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1152154.1152174"},{"key":"bibr14-1094342013507960","unstructured":"HPC Challenge Benchmarks (2005) Web resource. Available at: http:\/\/icl.cs.utk.edu\/hpcc\/ (accessed 30 September 2013)."},{"key":"bibr15-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/ISPASS.2012.6189201"},{"key":"bibr16-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2006.85"},{"key":"bibr17-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2009.5306801"},{"key":"bibr18-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1735688.1735696"},{"key":"bibr19-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2008.31"},{"key":"bibr20-1094342013507960","unstructured":"LINPACK (1979) Web resource. Available at: http:\/\/www.netlib.org\/linpack\/ (accessed 30 September 2013)."},{"key":"bibr21-1094342013507960","unstructured":"MATLAB Statistics Toolbox (2013) Web resource. Available at: http:\/\/www.mathworks.com (accessed 30 September 2013)."},{"key":"bibr22-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/2063384.2063402"},{"key":"bibr23-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1815961.1815992"},{"key":"bibr24-1094342013507960","unstructured":"NVIDIA CUDA Programming Guide (2013) Web resource. Available at: http:\/\/developer.nvidia.com\/object\/gpucomputing.html (accessed 30 September 2013)."},{"key":"bibr25-1094342013507960","unstructured":"NVIDIA Profiler User\u2019s Guide (2013) Web resource. Available at: http:\/\/docs.nvidia.com\/cuda\/profiler-users-guide\/index.html (accessed 30 September 2013)."},{"key":"bibr26-1094342013507960","unstructured":"OpenCL (2013) Web resource. Available at: http:\/\/www.khronos.org\/opencl\/ (accessed 30 September 2013)."},{"key":"bibr27-1094342013507960","unstructured":"Parboil Benchmark Suite (2012) Web resource. Available at: http:\/\/impact.crhc.illinois.edu\/Parboil\/parboil.aspx (accessed 30 September 2013)."},{"key":"bibr28-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1250662.1250713"},{"key":"bibr29-1094342013507960","volume-title":"Measuring program similarity for efficient benchmarking and performance analysis of computer systems","author":"Phansalkar AS","year":"2007"},{"key":"bibr30-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC.2011.6114192"},{"key":"bibr31-1094342013507960","doi-asserted-by":"crossref","unstructured":"Snavely A, Carrington L, Wolter N, (2002) A framework for performance modeling and prediction. In:\n                      Proceedings of the 2002 ACM\/IEEE conference on supercomputing\n                      .","DOI":"10.1109\/SC.2002.10004"},{"key":"bibr32-1094342013507960","unstructured":"SPEC Accelerator Benchmark Suite (2011) Web resource. Available at: http:\/\/www.spec.org\/hpg\/press\/acceleratorbmk.html."},{"key":"bibr33-1094342013507960","unstructured":"STREAM (2007) Web resource. Available at: http:\/\/www.cs.virginia.edu\/stream\/ (accessed 30 September 2013)."},{"key":"bibr34-1094342013507960","unstructured":"The Standard Performance Evaluation Corporation (SPEC) (2013) Web resource. Available at: http:\/\/www.spec.org."},{"key":"bibr35-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/ISPASS.2010.5452013"},{"key":"bibr36-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/223982.223990"},{"key":"bibr37-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1145\/1950365.1950408"},{"key":"bibr38-1094342013507960","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA.2011.5749745"}],"container-title":["The International Journal of High Performance Computing Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/1094342013507960","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/1094342013507960","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/1094342013507960","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T08:19:18Z","timestamp":1777450758000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/1094342013507960"}},"subtitle":["Correlating the performance of GPU benchmarks"],"short-title":[],"issued":{"date-parts":[[2013,10,11]]},"references-count":38,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2014,5]]}},"alternative-id":["10.1177\/1094342013507960"],"URL":"https:\/\/doi.org\/10.1177\/1094342013507960","relation":{},"ISSN":["1094-3420","1741-2846"],"issn-type":[{"value":"1094-3420","type":"print"},{"value":"1741-2846","type":"electronic"}],"subject":[],"published":{"date-parts":[[2013,10,11]]}}}