BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Europe/Zurich BEGIN:DAYLIGHT DTSTART:20240331T030000 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZNAME:CEST END:DAYLIGHT BEGIN:STANDARD DTSTART:20241027T020000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:CET END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20240910T073630Z UID:491C031B-5495-477F-B006-33CD71ADAA3A DTSTART;TZID=Europe/Zurich:20240909T140000 DTEND;TZID=Europe/Zurich:20240909T150000 DESCRIPTION:Speaker: Dr. Kwabena Boahen\n\nAffiliation: Stanford University \n\nTitle: Scaling Knowledge Processing from 2D Chips to 3D Brains\n\nAbst ract: Artificial intelligence (AI) now advances by performing twice as man y multiplications every two months\, but the semiconductor industry tiles twice as many multipliers on a chip every two years. Moreover\, the return s from tiling these multipliers ever more densely now diminish because sig nals must travel relatively farther and farther. Thus\, communicating now consumes much more energy and generates much more heat than computing does . Although travel can be shortened by stacking multipliers tiled in two di mensions in the third dimension\, such a solution acutely reduces the avai lable surface area for dissipating heat. My recent reconception of the bra in’s fundamental unit of computation cuts communication by moving away f rom synaptocentric learning to dendrocentric learning. Synaptocentric lear ning weights input signals precisely across an entire arbor of dendrite to discriminate spatial patterns of input. Current AI\, by using dot-product s to emulate synaptic weighting\, realizes this 60-year-old conception but produces dense output signals. Dendrocentric learning orders input signal s meticulously along a short stretch of dendrite to detect a particular sp atiotemporal pattern of inputs. I will illustrate how dendrocentric learni ng AI\, by using a string of ferroelectric transistors to emulate a stretc h of dendrite\, could signal sparsely and thus enable knowledge processing to scale from 2D chips to 3D brains.\n\nBiography: Kwabena Boahen is a Pr ofessor of Bioengineering\, Electrical Engineering\, and by courtesy Compu ter Science at Stanford University\; an investigator in Stanford’s Bio-X Institute\, System X Alliance\, and Wu Tsai Neurosciences Institute\; and the founding director of Stanford’s Brains in Silicon Lab. His group mo dels the nervous system computationally to elucidate principles of neural design at the cellular\, circuit\, and systems levels\; and synthesizes ne uromorphic electronic systems that scale energy-use with size as efficient ly as the brain does. He earned a doctorate in Computation and Neural Syst ems at the California Institute of Technology in 1997. From 1997 to 2005 h e was on the faculty of University of Pennsylvania\, Philadelphia PA. His research has resulted in over a hundred publications\, including a cover s tory in Scientific American featuring his lab’s work on a silicon retina and a silicon tectum that “wire together” automatically (May 2005). H e has received several distinguished honors\, including a Packard Fellowsh ip for Science and Engineering (1999) and a National Institutes of Health Director’s Pioneer Award (2006). He was elected a fellow of the American Institute for Medical and Biological Engineering (2016) and of the Instit ute of Electrical and Electronic Engineers (2016) in recognition of his la b’s work on Neurogrid\, an iPad-size platform that emulates the cerebral cortex in biophysical detail and at functional scale\, a combination that hitherto required a supercomputer. He has trained over twenty graduate st udents and mentored four postdoctoral researchers\, including the designer s of IBM’s TrueNorth chip and NeuraLink’s implantable chip and the fou nders of Femtosense and Dexterity\, two Silicon Valley start-ups.\n\nCo-sp onsored by: Institute of Neuroinformatics\, UZH and ETHZ\n\nAgenda: \n13:5 0 Welcome\n\n14:00 - 15:15 Lecture and Q&A\n\nRoom: F51\, Bldg: 35\, Unive rsity of Zurich\, Winterthurerstrasse 190\, Zurich\, Switzerland\, Switzer land\, 8057 LOCATION:Room: F51\, Bldg: 35\, University of Zurich\, Winterthurerstrasse 190\, Zurich\, Switzerland\, Switzerland\, 8057 ORGANIZER:shih@ini.uzh.ch SEQUENCE:12 SUMMARY:IEEE Swiss CAS/ED Sponsored Lecture by Dr. Kwabena Boahen URL;VALUE=URI:https://events.vtools.ieee.org/m/430578 X-ALT-DESC:Description: <br /><p><strong>Speaker: </strong>Dr. Kwabena Boah en&nbsp\;</p>\n<p><strong>Affiliation:&nbsp\;</strong>Stanford University< /p>\n<p><strong>Title: </strong>Scaling Knowledge Processing from 2D Chips to 3D Brains&nbsp\;</p>\n<p><strong>Abstract: </strong>Artificial intelli gence (AI) now advances by performing twice as many multiplications every two months\, but the semiconductor industry tiles twice as many multiplier s on a chip every two years. Moreover\, the returns from tiling these mult ipliers ever more densely now diminish because signals must travel relativ ely farther and farther. Thus\, communicating now consumes much more energ y and generates much more heat than computing does. Although travel can be shortened by stacking multipliers tiled in two dimensions in the third di mension\, such a solution acutely reduces the available surface area for d issipating heat. My recent reconception of the brain&rsquo\;s fundamental unit of computation cuts communication by moving away from synaptocentric learning to dendrocentric learning. Synaptocentric learning weights input signals precisely across an entire arbor of dendrite to discriminate spati al patterns of input. Current AI\, by using dot-products to emulate synapt ic weighting\, realizes this 60-year-old conception but produces dense out put signals. Dendrocentric learning orders input signals meticulously alon g a short stretch of dendrite to detect a particular spatiotemporal patter n of inputs. I will illustrate how dendrocentric learning AI\, by using a string of ferroelectric transistors to emulate a stretch of dendrite\, cou ld signal sparsely and thus enable knowledge processing to scale from 2D c hips to 3D brains.</p>\n<p><strong>Biography: </strong>Kwabena Boahen is a Professor of Bioengineering\, Electrical Engineering\, and by courtesy Co mputer Science at Stanford University\; an investigator in Stanford&rsquo\ ;s Bio-X Institute\, System X Alliance\, and Wu Tsai Neurosciences Institu te\; and the founding director of Stanford&rsquo\;s Brains in Silicon Lab. His group models the nervous system computationally to elucidate principl es of neural design at the cellular\, circuit\, and systems levels\; and s ynthesizes neuromorphic electronic systems that scale energy-use with size as efficiently as the brain does. He earned a doctorate in Computation an d Neural Systems at the California Institute of Technology in 1997. From 1 997 to 2005 he was on the faculty of University of Pennsylvania\, Philadel phia PA. His research has resulted in over a hundred publications\, includ ing a cover story in Scientific American featuring his lab&rsquo\;s work o n a silicon retina and a silicon tectum that &ldquo\;wire together&rdquo\; automatically (May 2005). He has received several distinguished honors\, including a Packard Fellowship for Science and Engineering (1999) and a Na tional Institutes of Health Director&rsquo\;s Pioneer Award (2006). He was elected a fellow of the American Institute for Medical and Biological Eng ineering (2016) and of the Institute of Electrical and Electronic Engineer s (2016) in recognition of his lab&rsquo\;s work on Neurogrid\, an iPad-si ze platform that emulates the cerebral cortex in biophysical detail and at functional scale\, a combination that hitherto required a supercomputer. He has trained over twenty graduate students and mentored four postdoctora l researchers\, including the designers of IBM&rsquo\;s TrueNorth chip and NeuraLink&rsquo\;s implantable chip and the founders of Femtosense and De xterity\, two Silicon Valley start-ups.</p><br /><br />Agenda: <br /><p>13 :50 Welcome</p>\n<p>14:00 - 15:15 Lecture and Q&amp\;A</p>\n<p>&nbsp\;</p> END:VEVENT END:VCALENDAR