Dynamic scheduling and varying decomposition granularity are well-known techniques for achieving high performance in parallel computing. Heterogeneous clusters with highly data-parallel processors, such as GPUs, present unique problems for the application of these techniques. These systems reveal a dichotomy between grain sizes: decompositions ideal for the CPUs may yield insufficient data-parallelism for accelerators, and decompositions targeted at the GPU may decrease performance on the CPU. This problem is typically ameliorated by statically scheduling a fixed amount of work for agglomeration. However, determining the ideal amount of work to compose requires experimentation because it varies between architectures and problem configurations. This paper describes a novel methodology for dynamically agglomerating work units at runtime and scheduling them on accelerators. This approach is demonstrated in the context of two applications: an n-body particle simulation, which offloads particle interaction work; and a parallel dense LU solver, which relocates DGEMM kernels to the GPU. In both cases dynamic agglomeration yields comparable or better results over statically scheduling the work across a variety of system configurations.<\/p>\n","protected":false},"excerpt":{"rendered":"
Dynamic scheduling and varying decomposition granularity are well-known techniques for achieving high performance in parallel computing. Heterogeneous clusters with highly data-parallel processors, such as GPUs, present unique problems for the application of these techniques. These systems reveal a dichotomy between grain sizes: decompositions ideal for the CPUs may yield insufficient data-parallelism for accelerators, and decompositions […]<\/p>\n","protected":false},"author":351,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[11,89,3],"tags":[1782,14,452,258,20,286,244],"class_list":["post-7274","post","type-post","status-publish","format-standard","hentry","category-computer-science","category-nvidia-cuda","category-paper","tag-computer-science","tag-cuda","tag-heterogeneous-systems","tag-n-body-simulation","tag-nvidia","tag-particle-simulation","tag-tesla-s1070"],"views":1993,"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/7274","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/users\/351"}],"replies":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=7274"}],"version-history":[{"count":0,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/7274\/revisions"}],"wp:attachment":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7274"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7274"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7274"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}