Large heterogeneous data centers of today lack methods to appraise the best fitting solutions regarding, among others, hardware acquisition cost, development time, and performance. Especially resource intensive applications benefit from increased data center utilization to leverage heterogeneous resources and accelerators. In this paper, we implement various methods to accelerate a seismic modeling application, which is available for CPU, GPU, and FPGA. With the underlying heterogeneous environment, the current programming standard OpenCL is examined regarding CPUs and GPUs, and compared to traditional acceleration approaches in order to evaluate sets of platforms. Based on the variety of available versions, a flow is introduced, which allows to catalog best solutions by experimenting with different implementations for available hardware platforms. We encourage to derive indicators as hints for data center operators with respect to finding a cost-benefit trade-off, which must also be observed over time. The results highlight the GPU and FPGA implementations, and correlate performance optimizations with development time, regarding the seismic application and the underlying hardware platforms.<\/p>\n","protected":false},"excerpt":{"rendered":"
Large heterogeneous data centers of today lack methods to appraise the best fitting solutions regarding, among others, hardware acquisition cost, development time, and performance. Especially resource intensive applications benefit from increased data center utilization to leverage heterogeneous resources and accelerators. In this paper, we implement various methods to accelerate a seismic modeling application, which is […]<\/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":[303,192,90,3],"tags":[1801,377,1798,452,20,1793,298,241,856,851],"class_list":["post-5688","post","type-post","status-publish","format-standard","hentry","category-earth-and-space-sciences","category-geoscience","category-opencl","category-paper","tag-earth-and-space-sciences","tag-fpga","tag-geoscience","tag-heterogeneous-systems","tag-nvidia","tag-opencl","tag-optimization","tag-seismic-modeling","tag-seismology","tag-tesla-t10p"],"views":2182,"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/5688","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=5688"}],"version-history":[{"count":0,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/5688\/revisions"}],"wp:attachment":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5688"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5688"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5688"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}