BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000712Z
LOCATION:405-406-407
DTSTART;TZID=America/Denver:20231117T110000
DTEND;TZID=America/Denver:20231117T113000
UID:submissions.supercomputing.org_SC23_sess462_ws_hmem103@linklings.com
SUMMARY:CachedArrays:  API and Framework to Optimize Data Movement for Het
 erogeneous Memory Systems
DESCRIPTION:Mark Hildebrand, Jason Lowe-Power, and Venkatesh Akella (Unive
 rsity of California, Davis)\n\nWe propose a new framework called CachedArr
 ays and a set of APIs to address the data tiering problem in large scale h
 eterogeneous and disaggregated memory systems. The proposed framework oper
 ates at a variable size object granularity and allows the programmer to sp
 ecify semantic hints about future use of data via a Policy API, which are 
 used by a Data Manager to choose when and where to place a particular data
  object using a data management API, thus bridging the semantic gap betwee
 n the programmer and the platform-specific hardware details, and optimizin
 g overall performance.  We evaluate the proposed framework on a real hardw
 are platform with terabytes of memory consisting of NVRAM and DRAM on larg
 e scale ML training workloads such CNNs, DNNs, and DLRM that exhibit diffe
 rent data access and usage patterns.\n\nTag: Data Movement and Memory, Fau
 lt Handling and Tolerance, Hardware Technologies, Heterogeneous Computing,
  I/O and File Systems, Performance Measurement, Modeling, and Tools, Progr
 amming Frameworks and System Software, Security\n\nRegistration Category: 
 Workshop Reg Pass\n\nSession Chairs: João Barreto (INESC-ID, IST, Universi
 ty of Lisbon; Universidade de Lisboa); Hatem ElShazly (Sony); Antonio J. P
 eña (Barcelona Supercomputing Center (BSC); Universitat Politècnica de Cat
 alunya (UPC), Spain); and Harald Servat (Intel Corporation)\n\n
END:VEVENT
END:VCALENDAR