BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240626T180033Z
LOCATION:3001\, 3rd Floor
DTSTART;TZID=America/Los_Angeles:20240625T153000
DTEND;TZID=America/Los_Angeles:20240625T154500
UID:dac_DAC 2024_sess105_RESEARCH1402@linklings.com
SUMMARY:Sharry：An Efficient and Sharing Far Memory System
DESCRIPTION:Research Manuscript\n\nChen Chen, Yuhang Huang, Shuiguang Deng
 , Jianwei Yin, and Xinkui Zhao (Zhejiang University)\n\nFar Memory System(
 FMS) allows applications to access memory on remote machines(called memory
  nodes). However, existing FMSs can`t deal with large loads and have low e
 fficiency in utilizing remote memory, which leads to the inability to shar
 e memory nodes among multiple processes, limiting the scalability of FMS. 
 \nIn this paper, we propose Sharry, an efficient Sharing FMS. Sharry manag
 es memory objects from multiple processes within a unified address space, 
 avoiding the overhead of space switching. Sharry also optimizes the utiliz
 ation of remote memory with fine-grained memory management. Additionally, 
 Sharry offloads memory allocation to dedicated CPU core in order to handle
  larger loads in the sharing scenario. \nCompared to state-of-the-art FMS,
  Sharry improves memory utilisation by 45%, causing only 9% performance de
 gradation when multiple processes sharing single memory node.\n\nTopic: AI
 \n\nKeyword: AI/ML Application and Infrastructure\n\nSession Chairs: Hongx
 iang Fan (Imperial College London; Samsung AI Center, UK) and Xiaoxuan Yan
 g (University of Virginia, Stanford University)
END:VEVENT
END:VCALENDAR