BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Asia/Jerusalem BEGIN:DAYLIGHT DTSTART:20250328T030000 TZOFFSETFROM:+0200 TZOFFSETTO:+0300 RRULE:FREQ=YEARLY;BYDAY=-1FR;BYMONTH=3 TZNAME:IDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20241027T010000 TZOFFSETFROM:+0300 TZOFFSETTO:+0200 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:IST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20241121T161453Z UID:4C09DB50-F8DE-4A9A-AA7B-88685B7A1635 DTSTART;TZID=Asia/Jerusalem:20241120T173000 DTEND;TZID=Asia/Jerusalem:20241120T190000 DESCRIPTION:Abstract:\n\nDeep learning can provide superior performance in many fields of applications. However\, the cost of implementing deep learn ing models in practical applications is expensive. Deep learning models ar e both computation intensive and memory intensive. Computation is an impor tant aspect for deep learning. It can determine the latency that is how fa st the results can be obtained. In this seminar\, computer arithmetic for deep learning will be discussed. This lecture will start with discussing t he computation requirements of deep learning models and layers. Then\, sev eral computer arithmetic designs for deep learning in the literature will be used as examples. Finally\, future trends of computer arithmetic for de ep learning computation will be discussed.\n\nPosit is designed as an alte rnative to IEEE 754 floating-point format for many applications. It has no n-uniformed number distribution\, and it can provide a much larger dynamic range than IEEE floating-point format. These make posit especially suitab le for deep learning applications. In recent years\, more and more posit b ased deep learning hardware accelerators appear in the literature. In this lecture\, the basics of posit format and the corresponding posit-based ar ithmetic units available in the literature\, including adder\, multiplier\ , multiply-accumulate unit\, and quire operator\, will be discussed. Then\ , several posit-based deep learning processors for deep learning inference and training will be discussed. Finally\, the trends and challenges of po sit arithmetic units and posit based deep learning processors will be disc ussed to motivate more related research works.\n\nDeep learning applicatio ns will be shared with the audience.\n\nBio:\n\nSeokbum Ko is currently a Professor at the Department of Electrical and Computer Engineering and the Division of Biomedical Engineering\, University of Saskatchewan\, Canada. He received his PhD from the University of Rhode Island\, USA in 2002.\n\ nHis areas of research interest include computer architecture/arithmetic\, efficient hardware implementation of compute-intensive applications\, dee p learning processor architecture and biomedical engineering.\n\nHe is an IEEE Cicuits and Systems Society Distinguished Lecturer (2024-2025)\, a se nior member of IEEE circuits and systems society and an associate editor f or IEEE TVLSI\, IEEE TCAS-II\, IEEE Access and IET Computers & Digital Tec hniques. He is an active member of IEEE CAS Technical Committee\, IEEE P31 09\, IEEE754-2029\, IEEE Domain-Specific Accelerators Standarads Committee and IEEE Emerging Processor Systems Standards Committee. He was an associ ate editor for IEEE TCASI (2019-2021).\n\nThe webinar is free but registra tion is required.\n\nZoom link will be sent after registration.\n\nVirtual : https://events.vtools.ieee.org/m/446950 LOCATION:Virtual: https://events.vtools.ieee.org/m/446950 ORGANIZER:shahar@ee.technion.ac.il SEQUENCE:127 SUMMARY:Efficient Hardware Implementation of Deep Learning Computation and its Application - Prof. Seokbum Ko\, University of Saskatchewan\, Canada URL;VALUE=URI:https://events.vtools.ieee.org/m/446950 X-ALT-DESC:Description: <br /><p><strong>Abstract:</strong></p>\n<p>Deep le arning can provide superior performance in many fields of applications. Ho wever\, the cost of implementing deep learning models in practical applica tions is expensive. Deep learning models are both computation intensive an d memory intensive. Computation is an important aspect for deep learning. It can determine the latency that is how fast the results can be obtained. In this seminar\, computer arithmetic for deep learning will be discussed . This lecture will start with discussing the computation requirements of deep learning models and layers. Then\, several computer arithmetic design s for deep learning in the literature will be used as examples. Finally\, future trends of computer arithmetic for deep learning computation will be discussed.</p>\n<p>Posit is designed as an alternative to IEEE 754 floati ng-point format for many applications. It has non-uniformed number distrib ution\, and it can provide a much larger dynamic range than IEEE floating- point format. These make posit especially suitable for deep learning appli cations. In recent years\, more and more posit based deep learning hardwar e accelerators appear in the literature. In this lecture\, the basics of p osit format and the corresponding posit-based arithmetic units available i n the literature\, including adder\, multiplier\, multiply-accumulate unit \, and quire operator\, will be discussed. Then\, several posit-based deep learning processors for deep learning inference and training will be disc ussed. Finally\, the trends and challenges of posit arithmetic units and p osit based deep learning processors will be discussed to motivate more rel ated research works.</p>\n<p>Deep learning applications will be shared wit h the audience.</p>\n<p><strong>Bio:&nbsp\;</strong></p>\n<p>Seokbum Ko is currently a Professor at the Department of Electrical and Computer Engine ering and the Division of Biomedical Engineering\, University of Saskatche wan\, Canada. He received his PhD from the University of Rhode Island\, US A in 2002.</p>\n<p>His areas of research interest include computer archite cture/arithmetic\, efficient hardware implementation of compute-intensive applications\, deep learning processor architecture and biomedical enginee ring.</p>\n<p>He is an IEEE Cicuits and Systems Society Distinguished Lect urer (2024-2025)\, a senior member of IEEE circuits and systems society an d an associate editor for IEEE TVLSI\, IEEE TCAS-II\, IEEE Access and IET Computers &amp\; Digital Techniques. He is an active member of IEEE CAS Te chnical Committee\, IEEE P3109\, IEEE754-2029\, IEEE Domain-Specific Accel erators Standarads Committee and IEEE Emerging Processor Systems Standards Committee. He was an associate editor for IEEE TCASI (2019-2021).</p>\n<p ><strong>The webinar is free but registration is required.</strong></p>\n< p><strong>Zoom link will be sent after registration.&nbsp\;</strong></p> END:VEVENT END:VCALENDAR