{"id":171367,"date":"2014-06-16T13:15:56","date_gmt":"2014-06-16T13:15:56","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/project\/compression-accelerators\/"},"modified":"2019-08-19T10:54:23","modified_gmt":"2019-08-19T17:54:23","slug":"compression-accelerators","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/compression-accelerators\/","title":{"rendered":"Compression Accelerators"},"content":{"rendered":"

Data compression is essential to large-scale data centers to save both storage and network bandwidth. Current software based method suffers from high computational cost with limited performance.
\nIn this project, we are migrating the fundamental workload of the computer system to FPGA accelerator, aiming high throughput performance and high energy efficiency, as well as freeing some CPU resources.<\/p>\n

\n

Target algorithm<\/strong><\/p>\n

Xpress Compression Algorithm (opens in new tab)<\/span><\/a>\u00a0is Microsoft compression format that combines the dictionary based LZ77 method and Huffman encoding, similar to popular GZIP compression. Xpress9 is an advanced branch of Xpress family\u00a0algorithms targeting higher compression ratio with more optimization on both stages.<\/p>\n

System Architecture<\/strong><\/p>\n

The system architecture supports up to 128 multi-threaded compression contexts with custom PCIe interface and queue managements.\u00a0We integrate up to 7 compression engines on Altera Stratix D5 FPGA\u00a0each of which\u00a0accelerates\u00a0full features of\u00a0Xpress9 algorithm.\u00a0The hardware scheduler maximizes throughput performance of the engines.<\/p>\n

\"hwcomp_arch2\"<\/p>\n

Results\u00a0&\u00a0Future Projection<\/strong><\/p>\n

The proposed hardware compressor achieved 6% better and 30x more throughput than software based\u00a0GZIP compression with level 9 (best) optimization on a single Zeon core. We are also targeting other compression domains such as low compression – high throughput, high compression – low throughput to push this Pareto curve of compression ratio vs throughput with hardware acceleration.<\/p>\n

\"hwcomp_proj\"<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

Data compression is essential to large-scale data centers to save both storage and network bandwidth. Current software based method suffers from high computational cost with limited performance. In this project, we are migrating the fundamental workload of the computer system to FPGA accelerator, aiming high throughput performance and high energy efficiency, as well as freeing […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"research-area":[13552,13547],"msr-locale":[268875],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-171367","msr-project","type-msr-project","status-publish","hentry","msr-research-area-hardware-devices","msr-research-area-systems-and-networking","msr-locale-en_us","msr-archive-status-active"],"msr_project_start":"2014-06-16","related-publications":[166686,166838,168381],"related-downloads":[],"related-videos":[],"related-groups":[144666],"related-events":[],"related-opportunities":[],"related-posts":[],"related-articles":[],"tab-content":[{"id":0,"name":"Collaborators","content":"

\r\n\r\nScott Hauck<\/a>, University of Washington as a visiting researcher\r\n\r\nJinwook Oh<\/a>, IBM TJ Watson Research Center\u00a0as an intern (Sep 2012)\r\n\r\nJanarbek Matai<\/a>, UC San Diego as an intern (June 2013)\r\n\r\n<\/div>"}],"slides":[],"related-researchers":[],"msr_research_lab":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/171367"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/171367\/revisions"}],"predecessor-version":[{"id":604266,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/171367\/revisions\/604266"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=171367"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=171367"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=171367"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=171367"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=171367"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}