{"id":509,"date":"2014-06-16T15:12:01","date_gmt":"2014-06-16T15:12:01","guid":{"rendered":"https:\/\/blogs.technet.microsoft.com\/inside_microsoft_research\/2014\/06\/16\/catapult-moving-beyond-cpus-in-the-cloud\/"},"modified":"2016-07-20T07:30:01","modified_gmt":"2016-07-20T14:30:01","slug":"catapult-moving-beyond-cpus-in-the-cloud","status":"publish","type":"post","link":"https:\/\/www.microsoft.com\/en-us\/research\/blog\/catapult-moving-beyond-cpus-in-the-cloud\/","title":{"rendered":"Catapult: Moving Beyond CPUs in the Cloud"},"content":{"rendered":"

Posted by Rob Knies<\/span><\/p>\n

\"Field-programmable (opens in new tab)<\/span><\/a><\/p>\n

Operating a datacenter at web scale requires managing many conflicting requirements. The ability to deliver computation at a high level and speed is a given, but because of the demands such a facility must meet, a datacenter also needs flexibility. Additionally, it must be efficient in its use of power, keeping costs as low as possible.<\/p>\n

Addressing often conflicting goals is a challenge, leading datacenter providers to seek constant performance and efficiency improvements and to evaluate the merits of general-purpose versus task-tuned alternatives\u2014particularly in an era in which Moore\u2019s Law is nearing an end, as some suggest.<\/p>\n

Microsoft researchers and colleagues from Bing (opens in new tab)<\/span><\/a> have been collaborating with others from industry and academia to examine datacenter hardware alternatives, and their work, a project known as Catapult (opens in new tab)<\/span><\/a>, was presented in Minneapolis on June 16 during the 41st International Symposium on Computer Architecture (opens in new tab)<\/span><\/a> (ISCA).<\/p>\n

Their paper, titled A Reconfigurable Fabric for Accelerating Large-Scale Datacenter Services<\/a><\/em>\" href=\"http:\/\/research.microsoft.com\/apps\/pubs\/default.aspx?id=212001\" target=\"_blank\">A Reconfigurable Fabric for Accelerating Large-Scale Datacenter Services (opens in new tab)<\/span><\/a><\/em>, describes an effort to combine programmable hardware and software that uses field-programmable gate arrays (opens in new tab)<\/span><\/a> (FPGAs) to deliver performance improvements of as much as 95 percent.<\/p>\n

The significance of this work, says Peter Lee (opens in new tab)<\/span><\/a>, head of Microsoft Research, could be dramatic.<\/p>\n