{"id":369623,"date":"2017-03-07T16:11:10","date_gmt":"2017-03-08T00:11:10","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-item&p=369623"},"modified":"2018-10-16T21:41:54","modified_gmt":"2018-10-17T04:41:54","slug":"viyojit-decoupling-battery-dram-capacities-battery-backed-dram","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/viyojit-decoupling-battery-dram-capacities-battery-backed-dram\/","title":{"rendered":"Viyojit: Decoupling Battery and DRAM Capacities for Battery-Backed DRAM"},"content":{"rendered":"

Non-Volatile Memories (NVMs) can significantly improve the performance of data-intensive applications. A popular form of NVM is Battery-backed DRAM, which is available and in use today with DRAMs latency and without the endurance problems of emerging\u00a0NVM technologies. Modern servers can be provisioned with up-to 4 TB of DRAM, and provisioning battery backup to write out such large memories is hard because of the large battery sizes and the\u00a0added hardware and cooling costs. We present Viyojit, a system\u00a0that exploits the skew in write working sets of applications to provision substantially smaller batteries while still ensuring durability\u00a0for the entire DRAM capacity. Viyojit achieves this by bounding the\u00a0number of dirty pages in DRAM based on the provisioned battery\u00a0capacity and proactively writing out infrequently written pages to\u00a0an SSD. Even for write-heavy workloads with less skew than we\u00a0observe in analysis of real data center traces, Viyojit reduces the\u00a0required battery capacity to 11% of the original size, with a performance overhead of 7-25%. Thus, Viyojit frees battery-backed\u00a0DRAM from stunted growth of battery capacities and enables servers\u00a0with terabytes of battery-backed DRAM.<\/p>\n","protected":false},"excerpt":{"rendered":"

Non-Volatile Memories (NVMs) can significantly improve the performance of data-intensive applications. A popular form of NVM is Battery-backed DRAM, which is available and in use today with DRAMs latency and without the endurance problems of emerging\u00a0NVM technologies. Modern servers can be provisioned with up-to 4 TB of DRAM, and provisioning battery backup to write out […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13547],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-369623","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"","msr_edition":"International Symposium on Computer Architecture (ISCA'17)","msr_affiliation":"","msr_published_date":"2017-06-24","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"380771","msr_publicationurl":"http:\/\/isca17.ece.utoronto.ca\/doku.php","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"ISCA17_Viyojit","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/03\/ISCA17_Viyojit.pdf","id":380771,"label_id":0},{"type":"url","title":"http:\/\/isca17.ece.utoronto.ca\/doku.php","viewUrl":false,"id":false,"label_id":0}],"msr_related_uploader":"","msr_attachments":[{"id":0,"url":"http:\/\/isca17.ece.utoronto.ca\/doku.php"}],"msr-author-ordering":[{"type":"text","value":"Rajat Kateja","user_id":0,"rest_url":false},{"type":"user_nicename","value":"anbadam","user_id":31008,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=anbadam"},{"type":"text","value":"Sriram Govindan","user_id":0,"rest_url":false},{"type":"text","value":"Bikash Sharma","user_id":0,"rest_url":false},{"type":"text","value":"Greg Ganger","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[144927],"msr_project":[171456],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/369623"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":3,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/369623\/revisions"}],"predecessor-version":[{"id":538205,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/369623\/revisions\/538205"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=369623"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=369623"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=369623"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=369623"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=369623"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=369623"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=369623"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=369623"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=369623"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=369623"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=369623"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=369623"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=369623"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=369623"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=369623"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}