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:20260422T000711Z
LOCATION:501-502
DTSTART;TZID=America/Denver:20231115T163000
DTEND;TZID=America/Denver:20231115T170000
UID:submissions.supercomputing.org_SC23_sess297_gb109@linklings.com
SUMMARY:Scaling the “Memory Wall” for Multi-Dimensional Seismic Processing
  with Algebraic Compression on Cerebras CS-2 Systems
DESCRIPTION:David Keyes, Hatem Ltaief, and Yuxi Hong (King Abdullah Univer
 sity of Science and Technology (KAUST)); Leighton Wilson and Mathias Jacqu
 elin (Cerebras Systems); and Matteo Ravasi (King Abdullah University of Sc
 ience and Technology (KAUST))\n\nWe exploit the high memory bandwidth of A
 I-customized Cerebras CS-2 systems for seismic processing. Through low-ran
 k matrix approximation, memory hungry seismic applications fit onto memory
 -austere SRAM waferscale hardware, addressing a challenge arising in many 
 wave-equation-based algorithms that rely on multi-dimensional convolution 
 operators. Exploiting sparsity inherent in seismic data in the frequency d
 omain, we implement embarrassingly parallel tile low-rank matrix-vector mu
 ltiplications (TLR-MVM), which account for most of the elapsed time in MDC
  operations, to solve the Multi-Dimensional Deconvolution (MDD) inverse pr
 oblem. By reducing memory footprint along with arithmetic complexity, we f
 it a standard seismic benchmark dataset into the local memories of Cerebra
 s processing elements. TLR-MVM on 48 CS-2 systems in support of MDD gives 
 a sustained memory bandwidth of 92.58PB/s on 35,784,000 processing element
 s, a significant milestone that highlights the capabilities of AI-customiz
 ed architectures to enable a new generation of seismic algorithms that wil
 l empower multiple technologies of our low-carbon future.\n\nRegistration 
 Category: Tech Program Reg Pass\n\nSession Chair: Barbara Chapman (Hewlett
  Packard Enterprise (HPE), Stony Brook University)\n\n
END:VEVENT
END:VCALENDAR