BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Denver
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:MDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T010000
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:MST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250417T140204Z
UID:101B65AC-B645-434F-A605-119AD0C7297F
DTSTART;TZID=America/Denver:20250416T180000
DTEND;TZID=America/Denver:20250416T200000
DESCRIPTION:Abstract-\n\nThe massive data deluge from mobile\, IoT\, and ed
 ge devices\, together with powerful innovations in data science and hardwa
 re processing\, have established artificial intelligence (AI) as the corne
 rstone of modern medical\, automotive\, industrial automation\, and consum
 er electronics domains. Domain-specific AI accelerators now dominate CPUs 
 and GPUs for energy-efficient AI and machine learning processing. However\
 , the evolution of these electronic accelerators is facing fundamental lim
 its due to the slowdown of Moore’s law and the reliance on metal wires\,
  which severely bottleneck computational performance today. In this talk\,
  I will present my vision of how silicon photonics can drive an entirely n
 ew class of light-driven AI hardware accelerators that can provide orders 
 of magnitude energy improvements over today’s accelerators. I will discu
 ss the evolution of silicon photonics\, from integrated optics to photonic
  devices that can now be fabricated with low-cost CMOS-compatible manufact
 uring techniques. I will cover new directions in the design of robust and 
 secure photonic substrates for communication\, computation\, and storage t
 o support emerging AI applications based on\nLLMs\, graph processing\, and
  generative modeling. I will share experiences from my journey over the pa
 st two decades towards realizing viable silicon photonic architectures. I 
 will end the talk with a discussion of the open challenges to achieve unpa
 ralleled energy-efficiency and performance gains in future computing platf
 orms with silicon photonics.\n\nSudeep Pasricha\n\nSudeep Pasricha is a Ar
 am and Helga Budak Endowed Professor in the [Department of Electrical and 
 Computer Engineering\,](http://www.engr.colostate.edu/ece/)the [Department
  of Computer Science](http://www.cs.colostate.edu/)\, and the [Department 
 of Systems Engineering](https://www.engr.colostate.edu/se/) at [Colorado S
 tate University](http://www.colostate.edu/). He is currently Director of t
 he Embedded\, High Performance\, and Intelligent Computing ([EPIC](https:/
 /www.engr.colostate.edu/~sudeep/wp-content/uploads/epic_lab_poster.pdf)) L
 aboratory and the Chair of Computer Engineering. He is a former University
  Distinguished Monfort Professor and College of Engineering Rockwell-Ander
 son Endowed Professor.\n\n------------------------------------------------
 ---------------\n\nComing up In May-\n\nSpeaker(s): Sudeep\, \n\nAgenda: \
 n6:00 pm Doors Open\n\n6:30 pm Welcome-Kris Waage\n\n6:45 pm Did&#39;ja Hear? 
 Scott Evans\n\n7:00 pm Main Presentation\n\n8:30 pm End\n\nRoom: E105\, Bl
 dg: Engineering Building\, Colorado State University\, Isotope Drive\, For
 t Collins\, Colorado\, United States\, 80525\, Virtual: https://events.vto
 ols.ieee.org/m/474065
LOCATION:Room: E105\, Bldg: Engineering Building\, Colorado State Universit
 y\, Isotope Drive\, Fort Collins\, Colorado\, United States\, 80525\, Virt
 ual: https://events.vtools.ieee.org/m/474065
ORGANIZER:rtoftness@gmail.com
SEQUENCE:24
SUMMARY:From Electrons to Photons: The Dawn of Light-Based Computing for AI
URL;VALUE=URI:https://events.vtools.ieee.org/m/474065
X-ALT-DESC:Description: &lt;br /&gt;&lt;table class=&quot;fusionResponsiveContent cke_sho
 w_border&quot; cellspacing=&quot;10&quot; align=&quot;left&quot;&gt;\n&lt;tbody&gt;\n&lt;tr&gt;\n&lt;td class=&quot;fusion
 ResponsiveColumn&quot; valign=&quot;top&quot; data-fusion-class=&quot;&quot;&gt;\n&lt;p&gt;&lt;strong&gt;Abstract-
 &lt;/strong&gt;&lt;/p&gt;\n&lt;p&gt;The massive data deluge from mobile\, IoT\, and edge dev
 ices\, together with powerful innovations in data science and hardware pro
 cessing\, have established artificial intelligence (AI) as the cornerstone
  of modern medical\, automotive\, industrial automation\, and consumer ele
 ctronics domains. Domain-specific AI accelerators now dominate CPUs and GP
 Us for energy-efficient AI and machine learning processing. However\, the 
 evolution of these electronic accelerators is facing fundamental limits du
 e to the slowdown of Moore&amp;rsquo\;s law and the reliance on metal wires\, 
 which severely bottleneck computational performance today. In this talk\, 
 I will present my vision of how silicon photonics can drive an entirely ne
 w class of light-driven AI hardware accelerators that can provide orders o
 f magnitude energy improvements over today&amp;rsquo\;s accelerators. I will d
 iscuss the evolution of silicon photonics\, from integrated optics to phot
 onic devices that can now be fabricated with low-cost CMOS-compatible manu
 facturing techniques. I will cover new directions in the design of robust 
 and secure photonic substrates for communication\, computation\, and stora
 ge to support emerging AI applications based on&lt;br&gt;LLMs\, graph processing
 \, and generative modeling. I will share experiences from my journey over 
 the past two decades towards realizing viable silicon photonic architectur
 es. I will end the talk with a discussion of the open challenges to achiev
 e unparalleled energy-efficiency and performance gains in future computing
  platforms with silicon photonics.&lt;/p&gt;\n&lt;p&gt;&lt;strong&gt;Sudeep Pasricha&lt;/strong
 &gt;&lt;/p&gt;\n&lt;p&gt;&lt;strong&gt;&amp;nbsp\;&lt;/strong&gt;Sudeep Pasricha is a Aram and Helga Buda
 k Endowed Professor in the&amp;nbsp\;&lt;a href=&quot;http://www.engr.colostate.edu/ec
 e/&quot; data-cke-saved-href=&quot;http://www.engr.colostate.edu/ece/&quot;&gt;Department of
  Electrical and Computer Engineering\,&amp;nbsp\;&lt;/a&gt;the&amp;nbsp\;&lt;a href=&quot;http:/
 /www.cs.colostate.edu/&quot; data-cke-saved-href=&quot;http://www.cs.colostate.edu/&quot;
 &gt;Department of Computer Science&lt;/a&gt;\, and the&amp;nbsp\;&lt;a href=&quot;https://www.e
 ngr.colostate.edu/se/&quot; data-cke-saved-href=&quot;https://www.engr.colostate.edu
 /se/&quot;&gt;Department of Systems Engineering&lt;/a&gt;&amp;nbsp\;at&amp;nbsp\;&lt;a href=&quot;http:/
 /www.colostate.edu/&quot; data-cke-saved-href=&quot;http://www.colostate.edu/&quot;&gt;Color
 ado State University&lt;/a&gt;. He is currently Director of the Embedded\, High 
 Performance\, and Intelligent Computing (&lt;a href=&quot;https://www.engr.colosta
 te.edu/~sudeep/wp-content/uploads/epic_lab_poster.pdf&quot; data-cke-saved-href
 =&quot;https://www.engr.colostate.edu/~sudeep/wp-content/uploads/epic_lab_poste
 r.pdf&quot;&gt;EPIC&lt;/a&gt;) Laboratory and&amp;nbsp\;the Chair of Computer Engineering. H
 e is a former University Distinguished Monfort Professor and College of En
 gineering Rockwell-Anderson Endowed Professor.&lt;/p&gt;\n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;\n&lt;hr&gt;\n
 &lt;p&gt;&lt;strong&gt;Coming up In May-&lt;/strong&gt;&lt;/p&gt;\n&lt;p&gt;&amp;nbsp\;&lt;/p&gt;\n&lt;/td&gt;\n&lt;/tr&gt;\n&lt;
 /tbody&gt;\n&lt;/table&gt;&lt;br /&gt;&lt;br /&gt;Agenda: &lt;br /&gt;&lt;p&gt;6:00 pm Doors Open&lt;/p&gt;\n&lt;p&gt;6
 :30 pm Welcome-Kris Waage&lt;/p&gt;\n&lt;p&gt;6:45 pm Did&#39;ja Hear? Scott Evans&lt;/p&gt;\n&lt;p
 &gt;7:00 pm Main Presentation&amp;nbsp\;&lt;/p&gt;\n&lt;p&gt;8:30 pm&amp;nbsp\; End&lt;/p&gt;
END:VEVENT
END:VCALENDAR