{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T04:20:48Z","timestamp":1750306848195,"version":"3.41.0"},"reference-count":34,"publisher":"Association for Computing Machinery (ACM)","issue":"4","license":[{"start":{"date-parts":[[2013,12,1]],"date-time":"2013-12-01T00:00:00Z","timestamp":1385856000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Archit. Code Optim."],"published-print":{"date-parts":[[2013,12]]},"abstract":"<jats:p>Multi-GPU machines are being increasingly used in high-performance computing. Each GPU in such a machine has its own memory and does not share the address space either with the host CPU or other GPUs. Hence, applications utilizing multiple GPUs have to manually allocate and manage data on each GPU. Existing works that propose to automate data allocations for GPUs have limitations and inefficiencies in terms of allocation sizes, exploiting reuse, transfer costs, and scalability.<\/jats:p>\n          <jats:p>\n            We propose a scalable and fully automatic data allocation and buffer management scheme for affine loop nests on multi-GPU machines. We call it the Bounding-Box-based Memory Manager (BBMM). BBMM can perform at\n            <jats:italic>runtime<\/jats:italic>\n            , during standard set operations like union, intersection, and difference, finding subset and superset relations on hyperrectangular regions of array data (bounding boxes). It uses these operations along with some compiler assistance to identify, allocate, and manage data required by applications in terms of\n            <jats:italic>disjoint<\/jats:italic>\n            bounding boxes. This allows it to (1) allocate exactly or nearly as much data as is required by computations running on each GPU, (2) efficiently track buffer allocations and hence maximize data reuse across tiles and minimize data transfer overhead, and (3) and as a result, maximize utilization of the combined memory on multi-GPU machines. BBMM can work with any choice of parallelizing transformations, computation placement, and scheduling schemes, whether static or dynamic. Experiments run on a four-GPU machine with various scientific programs showed that BBMM reduces data allocations on each GPU by up to 75% compared to current allocation schemes, yields performance of at least 88% of manually written code, and allows excellent weak scaling.\n          <\/jats:p>","DOI":"10.1145\/2544100","type":"journal-article","created":{"date-parts":[[2014,1,6]],"date-time":"2014-01-06T20:42:39Z","timestamp":1389040959000},"page":"1-26","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":12,"title":["Automatic data allocation and buffer management for multi-GPU machines"],"prefix":"10.1145","volume":"10","author":[{"given":"Thejas","family":"Ramashekar","sequence":"first","affiliation":[{"name":"Indian Institute of Science, Karnataka, India"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Uday","family":"Bondhugula","sequence":"additional","affiliation":[{"name":"Indian Institute of Science, Karnataka, India"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2013,12]]},"reference":[{"key":"e_1_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.1145\/277650.277721"},{"key":"e_1_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.1145\/155090.155102"},{"key":"e_1_2_1_3_1","doi-asserted-by":"publisher","DOI":"10.1002\/cpe.1631"},{"key":"e_1_2_1_4_1","doi-asserted-by":"publisher","DOI":"10.1145\/1345206.1345210"},{"key":"e_1_2_1_5_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-11970-5_14"},{"key":"e_1_2_1_6_1","volume-title":"Clan: The Chunky Loop Analyzer user guide","author":"Bastoul C.","year":"2005","unstructured":"Bastoul , C. 2005 . Clan: The Chunky Loop Analyzer user guide . http:\/\/icps.u-strasbg.fr\/&sim;bastoul\/development\/clan\/docs\/clan.pdf. Bastoul, C. 2005. Clan: The Chunky Loop Analyzer user guide. http:\/\/icps.u-strasbg.fr\/&sim;bastoul\/development\/clan\/docs\/clan.pdf."},{"key":"e_1_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.1145\/2503210.2503289"},{"key":"e_1_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1145\/1375581.1375595"},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1145\/1065944.1065948"},{"volume-title":"Proceedings of the International Parallel and Distributed Processing Symposium (IPDPS\u201906)","author":"Classen M.","key":"e_1_2_1_10_1","unstructured":"Classen , M. and Griebl , M . 2006. Automatic code generation for distributed memory architectures in the polytope model . In Proceedings of the International Parallel and Distributed Processing Symposium (IPDPS\u201906) . Classen, M. and Griebl, M. 2006. Automatic code generation for distributed memory architectures in the polytope model. In Proceedings of the International Parallel and Distributed Processing Symposium (IPDPS\u201906)."},{"key":"e_1_2_1_11_1","unstructured":"CUDA 2011. NVIDIA CUDA programming model. http:\/\/developer.nvidia.com\/object\/cuda.html.  CUDA 2011. NVIDIA CUDA programming model. http:\/\/developer.nvidia.com\/object\/cuda.html."},{"volume-title":"Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques (PACT\u201913)","author":"Dathathri R.","key":"e_1_2_1_12_1","unstructured":"Dathathri , R. , Reddy , C. , Ramashekar , T. , and Bondhugula , U . 2013. Generating efficient data movement code for heterogeneous architectures with distributed memory . In Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques (PACT\u201913) . Dathathri, R., Reddy, C., Ramashekar, T., and Bondhugula, U. 2013. Generating efficient data movement code for heterogeneous architectures with distributed memory. In Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques (PACT\u201913)."},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-00722-4_17"},{"volume-title":"Integer Set Library. Sven Verdoolaege","key":"e_1_2_1_14_1","unstructured":"ISL 2012. Integer Set Library. Sven Verdoolaege , An Integer Set Library for Program Analysis . ISL 2012. Integer Set Library. Sven Verdoolaege, An Integer Set Library for Program Analysis."},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1145\/2259016.2259038"},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1145\/1993498.1993516"},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1145\/1941553.1941591"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1145\/2145816.2145827"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.1109\/SC.2010.36"},{"key":"e_1_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.1145\/1504176.1504194"},{"key":"e_1_2_1_21_1","volume-title":"GPU Computing SDK","author":"NVIDIA","year":"2010","unstructured":"NVIDIA GPU Computing SDK 2010 . NVIDIA GPU Computing SDK. NVIDIA GPU Computing SDK 2010. NVIDIA GPU Computing SDK."},{"key":"e_1_2_1_22_1","unstructured":"OpenACC 2012. OpenACC Application Programming Interface.  OpenACC 2012. OpenACC Application Programming Interface."},{"key":"e_1_2_1_23_1","unstructured":"OpenCL 2011. Open Computing Language. http:\/\/www.khronos.org\/opencl\/.  OpenCL 2011. Open Computing Language. http:\/\/www.khronos.org\/opencl\/."},{"key":"e_1_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1145\/2370816.2370824"},{"key":"e_1_2_1_25_1","unstructured":"Par4All 2012. HPC open source project - Par4All. http:\/\/www.par4all.org.  Par4All 2012. HPC open source project - Par4All. http:\/\/www.par4all.org."},{"key":"e_1_2_1_26_1","unstructured":"PGI 2012. Portland Group Inc. http:\/\/www.pgroup.com\/.  PGI 2012. Portland Group Inc. http:\/\/www.pgroup.com\/."},{"key":"e_1_2_1_27_1","unstructured":"Polybench 2012. The polybench project.  Polybench 2012. The polybench project."},{"key":"e_1_2_1_28_1","unstructured":"Polylib 2010. PolyLib - A library of polyhedral functions.  Polylib 2010. PolyLib - A library of polyhedral functions."},{"key":"e_1_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.1145\/2312005.2312025"},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1145\/2304576.2304625"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1145\/1989493.1989508"},{"key":"e_1_2_1_32_1","unstructured":"Tavarageri S. Hartono A. Baskaran M. Pouchet L.-N. Ramanujam J. and Sadayappan P. 2013. Parametric Tiling of Affine Loop Nests. Technical Report.  Tavarageri S. Hartono A. Baskaran M. Pouchet L.-N. Ramanujam J. and Sadayappan P. 2013. Parametric Tiling of Affine Loop Nests. Technical Report."},{"key":"e_1_2_1_33_1","doi-asserted-by":"publisher","DOI":"10.1145\/2400682.2400713"},{"key":"e_1_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.1145\/1735688.1735697"}],"container-title":["ACM Transactions on Architecture and Code Optimization"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/2544100","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/2544100","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T08:10:15Z","timestamp":1750234215000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/2544100"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2013,12]]},"references-count":34,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2013,12]]}},"alternative-id":["10.1145\/2544100"],"URL":"https:\/\/doi.org\/10.1145\/2544100","relation":{},"ISSN":["1544-3566","1544-3973"],"issn-type":[{"type":"print","value":"1544-3566"},{"type":"electronic","value":"1544-3973"}],"subject":[],"published":{"date-parts":[[2013,12]]},"assertion":[{"value":"2013-06-01","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2013-11-01","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2013-12-01","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}