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:20211207T054745Z
LOCATION:263
DTSTART;TZID=America/Chicago:20211116T103000
DTEND;TZID=America/Chicago:20211116T110000
UID:submissions.supercomputing.org_SC21_sess262_exforum106@linklings.com
SUMMARY:The Role of Embedded AI/ML Cores in Hardware-Based Accelerators
DESCRIPTION:Exhibitor Forum\n\nThe Role of Embedded AI/ML Cores in Hardwar
 e-Based Accelerators\n\nSchweitzer\n\n“When all you have is a hammer, all 
 the world’s a nail.” For decades in computing, all we had was the CPU, so 
 every computational problem was viewed from this perspective. Eventually, 
 math co-processors came along to offload instruction cycle-intensive math 
 operations. This then led to offload solutions via graphical processing un
 its (GPUs), field-programmable gate arrays (FPGAs), and, more recently, ap
 plication-specific integrated circuits (ASICs). Specialized Artificial Int
 elligence (AI)/Machine Learning (MLP) Processing cores have also been deve
 loped in recent years, which run on GPUs, FPGAs and/or ASICs. These comput
 ational platforms specialize in specific HPC problem domains where their c
 omputational resources offer the best ratio of results to power consumptio
 n. For FPGAs, HPC problems like genomic analysis or video data can be up t
 o 100x faster than GPUs. Various types of machine learning map exceptional
 ly well into MLP cores, dramatically improve training time. This talk will
  examine each of these computational platforms, but it will focus on the m
 ost recent two platforms, ASICs and FPGAs with enhanced MLP cores, diving 
 into why these platforms, for specific problem domains, are so much better
  than earlier competing platforms.\n\nTag: Architectures\n\nRegistration C
 ategory: Tech Program Reg Pass, Exhibit Hall Only
END:VEVENT
END:VCALENDAR