BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230124T171521Z
LOCATION:C140-142
DTSTART;TZID=America/Chicago:20221118T111000
DTEND;TZID=America/Chicago:20221118T113000
UID:submissions.supercomputing.org_SC22_sess444_ws_hipar101@linklings.com
SUMMARY:Expressing Hierarchical Code Optimizations via MDH-Based Schedules
DESCRIPTION:Workshop\n\nExpressing Hierarchical Code Optimizations via MDH
 -Based Schedules\n\nRasch, Schulze, Gorlatch\n\nThe optimization processes
  of modern compilers are often guided by performance experts who express o
 ptimization via so-called scheduling languages.  We present our work-in-pr
 ogress results toward a novel compiler whose scheduling language is based 
 on the approach of Multi-Dimensional Homomorphisms (MDH). We argue that ou
 r MDH-based scheduling language enables a structured, hierarchical optimiz
 ation process, by offering scheduling commands that systematically de- and
  re-compose computations to/from the memory and core hierarchies of state-
 of-the-art architectures (GPUs, multi-core CPUs, etc). Thereby, a performa
 nce expert expresses hierarchical code optimizations in a concise and stru
 ctured way, contributing to a simplified code optimization process.  Our f
 irst experiments on NVIDIA GPU and Intel CPU show that our scheduling lang
 uage is capable of expressing as hierarchical code optimizations the optim
 ization decisions of the popular deep learning compiler TVM.  Our experime
 nts also confirm that via auto-tuning, we are able to achieve better perfo
 rmance than TVM on both architectures.\n\nSession Format: Recorded\n\nTag:
  Algorithms, Architectures, Compilers, Computational Science, Exascale Com
 puting, Heterogeneous Systems, Hierarchical Parallelism, Memory Systems, P
 arallel Programming Languages and Models, Parallel Programming Systems, Re
 source Management and Scheduling\n\nRegistration Category: Workshop Reg Pa
 ss
END:VEVENT
END:VCALENDAR