BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Denver
BEGIN:DAYLIGHT
DTSTART:20240310T030000
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:MDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20241103T010000
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:MST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20241103T160718Z
UID:4674A12C-5745-4E4C-8474-AA33CB147A85
DTSTART;TZID=America/Denver:20241031T120000
DTEND;TZID=America/Denver:20241031T133000
DESCRIPTION:Please join us at UCCS For Our 2nd Electronic Devices/Circuits 
 &amp; Systems Meeting of 2024. Hear information on the latest storage developm
 ents from the UCCS Microelectronics Labs and enjoy lunch provided by the P
 ikes Peak Section.\n\nA New Nonvolatile Switch for the Age of AI\n\nProf. 
 Carlos Araujo\, Distinguished Professor\, IEEE Fellow\n\nUniversity of Col
 orado-Colorado Springs ECE Department\n\nMemory and artificial synapses ar
 e the two areas that require major innovation for the next generation of A
 I engines. Today’s memory of choice for the highest performance AI syste
 ms in High Bandwidth Memory (HBM) providing incredible bandwidth in a smal
 l space. A Micron HBM3E has an energy efficiency quoted at 1E-12 J per swi
 tching operation. While good\, there is much concern about the energy cons
 umption of AI data centers\, some which may require a nuclear power plant 
 to supply their energy.\n\nIn this lecture\, I describe a non-volatile swi
 tch that has been made at UCCS and IMEC using the Mott Quantum Switch (MQS
 ) at 45 nm design rules with energy performance of 1E-18 J. It has the pot
 ential to scale to the most advanced feature size using nanosheets made us
 ing common gate oxide Hafnium Oxide (HfO2) with doping with Carbon. These 
 fast switches have excellent nonvolatile properties from 2 degrees Kelvin 
 to 200 degrees C with less than 2 ns switching times and require only 3-4 
 mask steps. This makes them ideal for low power nonvolatile storage in fut
 ure AI systems. For synapses used in neural nets\, these switches have pro
 grammable levels of resistivity in a deterministic way for programmable ne
 uron weights.\n\nSpeaker(s): Carlos Araujo\, \n\nRoom:  Tom Saponas Lounge
 \, Bldg: Engineering Building\, 2nd Floor\, 1420 Austin Bluffs Parkway\, C
 olorado Springs\, Colorado\, United States\, 80918
LOCATION:Room:  Tom Saponas Lounge\, Bldg: Engineering Building\, 2nd Floor
 \, 1420 Austin Bluffs Parkway\, Colorado Springs\, Colorado\, United State
 s\, 80918
ORGANIZER:tkalkur@uccs.edu
SEQUENCE:35
SUMMARY: A New Nonvolatile Switch for the Age of AI
URL;VALUE=URI:https://events.vtools.ieee.org/m/442642
X-ALT-DESC:Description: &lt;br /&gt;&lt;p&gt;&lt;span style=&quot;font-size: 14pt\;&quot;&gt;&lt;strong&gt;Pl
 ease join us at UCCS For Our 2nd Electronic Devices/Circuits &amp;amp\; System
 s Meeting of 2024. &amp;nbsp\;Hear information on the latest storage developme
 nts from the UCCS Microelectronics Labs and enjoy lunch provided by the Pi
 kes Peak Section.&lt;/strong&gt;&lt;/span&gt;&lt;/p&gt;\n&lt;p&gt;&lt;span style=&quot;font-size: 14pt\;&quot;&gt;
 &lt;strong&gt;A New Nonvolatile Switch for the Age of AI&lt;/strong&gt;&lt;/span&gt;&lt;/p&gt;\n&lt;p
 &gt;&lt;span style=&quot;font-size: 14pt\;&quot;&gt;Prof. Carlos Araujo\, Distinguished Profe
 ssor\, IEEE Fellow&lt;/span&gt;&lt;/p&gt;\n&lt;p&gt;&lt;span style=&quot;font-size: 14pt\;&quot;&gt;Universi
 ty of Colorado-Colorado Springs ECE Department&lt;/span&gt;&lt;/p&gt;\n&lt;p&gt;&lt;span style=
 &quot;font-size: 14pt\;&quot;&gt;Memory and artificial synapses are the two areas that 
 require major innovation for the next generation of AI engines.&lt;span class
 =&quot;Apple-converted-space&quot;&gt;&amp;nbsp\; &lt;/span&gt;Today&amp;rsquo\;s memory of choice fo
 r the highest performance AI systems in High Bandwidth Memory (HBM) provid
 ing incredible bandwidth in a small space.&lt;span class=&quot;Apple-converted-spa
 ce&quot;&gt;&amp;nbsp\; &lt;/span&gt;A Micron&lt;span class=&quot;Apple-converted-space&quot;&gt;&amp;nbsp\; &lt;/s
 pan&gt;HBM3E has an energy efficiency quoted at 1E-12 J per switching operati
 on.&lt;span class=&quot;Apple-converted-space&quot;&gt;&amp;nbsp\; &lt;/span&gt;While good\, there i
 s much concern about the energy consumption of AI data centers\, some whic
 h may require a nuclear power plant to supply their energy. &lt;span class=&quot;A
 pple-converted-space&quot;&gt;&amp;nbsp\;&lt;/span&gt;&lt;/span&gt;&lt;/p&gt;\n&lt;p&gt;&lt;span style=&quot;font-size
 : 14pt\;&quot;&gt;In this lecture\, I describe a non-volatile switch that has been
  made at UCCS and IMEC using the Mott Quantum Switch (MQS) at 45 nm design
  rules with energy performance of 1E-18 J.&lt;span class=&quot;Apple-converted-spa
 ce&quot;&gt;&amp;nbsp\; &lt;/span&gt;It has the potential to scale to the most advanced feat
 ure size using nanosheets made using common gate oxide Hafnium Oxide (HfO&lt;
 sub&gt;2&lt;/sub&gt;) with doping with Carbon.&lt;span class=&quot;Apple-converted-space&quot;&gt;&amp;
 nbsp\; &lt;/span&gt;These fast switches have excellent nonvolatile properties fr
 om 2 degrees Kelvin to 200 degrees C with less than 2 ns switching times a
 nd require only 3-4 mask steps.&lt;span class=&quot;Apple-converted-space&quot;&gt;&amp;nbsp\;
  &lt;/span&gt;This makes them ideal for low power nonvolatile storage in future 
 AI systems.&lt;span class=&quot;Apple-converted-space&quot;&gt;&amp;nbsp\; &lt;/span&gt;For synapses
  used in neural nets\, these switches have programmable levels of resistiv
 ity in a deterministic way for programmable neuron weights.&lt;/span&gt;&lt;/p&gt;
END:VEVENT
END:VCALENDAR