BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T143141Z
LOCATION:B309
DTSTART;TZID=America/New_York:20241119T143000
DTEND;TZID=America/New_York:20241119T150000
UID:submissions.supercomputing.org_SC24_sess403_pap457@linklings.com
SUMMARY:autoGEMM: Pushing the Limits of Irregular Matrix Multiplication on
  Arm Architectures
DESCRIPTION:Du Wu (Tokyo Institute of Technology, RIKEN); Jintao Meng (She
 nzhen Institute of Advanced Technology, Chinese Academy of Sciences); Wenx
 i Zhu and Minwen Deng (Tencent); Xiao Wang (Oak Ridge National Laboratory 
 (ORNL)); Tao Luo (Agency for Science, Technology and Research (A*STAR)); M
 ohamed Wahib (RIKEN); and Yanjie Wei (Shenzhen Institute of Advanced Techn
 ology, Chinese Academy of Sciences)\n\nThis paper presents an open-source 
 library that pushes the limits of performance portability for irregular Ge
 neral Matrix Multiplication (GEMM) computations on the widely-used Arm arc
 hitectures. autoGEMM generates optimized kernels for various hardware conf
 igurations by auto-combining fragments of auto-generated micro-kernels tha
 t employ hand-written optimizations to maximize computational efficiency. 
 We optimize the kernel pipeline by tuning the register reuse and the data 
 load/store overlapping. In addition, we use a dynamic tiling scheme to gen
 erate balanced tile shapes, based on the shapes of the matrices. We build 
 autoGEMM on top of the TVM framework where our dynamic tiling scheme prune
 s the search space for TVM to identify the optimal combination of paramete
 rs for code optimization. Evaluations on five different classes of Arm chi
 ps demonstrate the advantages of autoGEMM. For small matrices, autoGEMM ac
 hieves 98% of peak and up to 2.0x speedup over state-of-the-art libraries 
 such as LIBXSMM and LibShalom. autoGEMM is available at:https://github.com
 /wudu98/autoGEMM.\n\nTag: Accelerators, Compilers, Embedded and/or Reconfi
 gurable Systems, Linear Algebra, Performance Evaluation and/or Optimizatio
 n Tools\n\nRegistration Category: Tech Program Reg Pass\n\nSession Chair: 
 Keita Teranishi (Oak Ridge National Laboratory (ORNL))\n\n
END:VEVENT
END:VCALENDAR