Sun 25 Mar 2018 14:00 - 14:30 - Session 2: Maps and Migrations Chair(s): Robert Wisniewski

This paper introduces gMig, an open-source and practical GPU live migration solution for full virtualization. By taking advantage of the dirty pattern of GPU workloads, gMig presents the One-Shot Pre-Copy combined with the hashing based Software Dirty Page technique to achieve efficient GPU live migration. Particularly, we propose three approaches for gMig: 1) Dynamic Graphics Address Remapping, which parses and manipulates GPU commands to adjust the address mapping to adapt to a different environment after migration, 2) Software Dirty Page, which utilizes a hashing based approach to detect page modification, overcomes the commodity GPU’s hardware limitation, and speeds up the migration by only sending the dirtied pages, 3) One-Shot Pre-Copy, which greatly reduces the rounds of pre-copy of graphics memory. Our evaluation shows that gMig achieves GPU live migration with an average downtime of 302 ms on Windows and 119 ms on Linux. With the help of Software Dirty Page, the number of GPU pages transferred during the downtime is effectively reduced by 80.0%.

Sun 25 Mar
Times are displayed in time zone: (GMT-04:00) Eastern Time (US & Canada) change

14:00 - 15:30: Research Papers - Session 2: Maps and Migrations
Chair(s): Robert WisniewskiIntel Corporation
vee-2018-Research-Papers14:00 - 14:30
Jiacheng Ma, Xiao ZhengIntel Corporation, Yaozu DongIntel Asia-Pacific R&D Ltd, China, Wentai LiShanghai Jiao Tong University, Zhengwei QiShanghai Jiao Tong University, Bingsheng HeNational University of Singapore, Haibing GuanShanghai Jiao Tong University
vee-2018-Research-Papers14:30 - 15:00
vee-2018-Research-Papers15:00 - 15:30
Yu XuShanghai Jiao Tong University, Jianguo YaoShanghai Jiao Tong University, Yaozu DongIntel Asia-Pacific R&D Ltd, China, Kun TianIntel Corporation, Xiao ZhengIntel Corporation, Haibing GuanShanghai Jiao Tong University