BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Chicago
BEGIN:DAYLIGHT
DTSTART:20240310T030000
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:CDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20241103T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:CST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240605T224300Z
UID:CF6B3EC1-8957-4703-BE41-694A5A14E4C3
DTSTART;TZID=America/Chicago:20240603T100000
DTEND;TZID=America/Chicago:20240603T130000
DESCRIPTION:The IEEE Central Texas CASS &amp; SSCS JT. Chapter\,\n\nCircuits An
 d Systems Society Outreach Program\,\n\nSilicon Laboratories\n\nPresent:\n
 \nFeed Your Mind - Practical Aspects of Machine Learning Circuits and Syst
 ems\n\nWhat is behind the buzzwords Machine Learning\, Artificial Intellig
 ence? In this lecture Prof. H. Li and A. Sanyal will present practical asp
 ects applied to circuits and systems solutions for everyday&#39;s life.\n\nAI 
 Models for Edge Computing: Hardware-aware Optimizations for Efficiency\n\n
 Abstract:   As artificial intelligence (AI) transforms various industries\
 , state-of-the-art models have exploded in size and capability. The growth
  in AI model complexity is rapidly outstripping hardware evolution\, makin
 g the deployment of these models on edge devices remain challenging. To en
 able advanced AI locally\, models must be optimized for fitting into the h
 ardware constraints. In this presentation\, we will first discuss how comp
 uting hardware designs impact the effectiveness of commonly used AI model 
 optimizations for efficiency\, including techniques like quantization and 
 pruning. Additionally\, we will present several methods\, such as hardware
 -aware quantization and structured pruning\, to demonstrate the significan
 ce of software/hardware co-design. We will also demonstrate how these meth
 ods can be understood via a straightforward theoretical framework\, facili
 tating their seamless integration in practical applications and their stra
 ightforward extension to distributed edge computing. At the conclusion of 
 our presentation\, we will share our insights and vision for achieving eff
 icient and robust AI at the edge.\n\nHealth management using intelligent w
 earables with mixed-signal AI\n\nAbstract: As medical wearables become mor
 e widely adopted for at-home/early diagnosis/health surveillance\, the vol
 ume of data produced by these devices are expected to reach thousands of p
 etabytes/month. Transmitting this large volume of data over the cloud for 
 processing will potentially emerge as a communication bottleneck and incre
 ase latency of decisions. Transmitting naively all data generated by a wea
 rable medical device is also costly in terms of power/energy- transmitter 
 is usually the highest consumer of energy in a sensor (at least 10~20x mor
 e energy than sensing). Key to addressing this data deluge is to increase 
 capabilities of the wearable devices to process information locally and ha
 ve on-device inference capabilities\, such as through embedding AI capabil
 ities into the wearable device that will allow extraction of key informati
 on from the sensor data. There needs to be balance between what can be pro
 cessed locally on-device with low power/energy and how to optimally decide
  the volume of data communication from the device (to cloud as an example)
 . The barriers to this approach lie in the computational complexity of AI 
 algorithms that makes it challenging to fit AI models on wearables with li
 mited resources. Some of the answers might lie in going back to early days
  of signal processing in silicon – developing analog circuit techniques 
 for AI development which will require collaborative innovations in both AI
  model development and analog circuit design techniques. In this talk\, I 
 will present our research on developing analog AI circuits and their demon
 strations with patient data with use cases from cardiovascular health moni
 toring and sepsis onset detection.\n\nCo-sponsored by: Mikko Sojonen - Sil
 icon Labs\n\nSpeaker(s): Helen Li\, Arindam Sanyal\n\nAgenda: \nProgram:\n
 \n- Introduction Dr. S. Pietri (NXP) and J. Elenes (Silabs)= 10 – 10:15 
 am\n- Dr. Hai &quot;Helen&quot; Li (Duke University) lecture= 10-11:15 am\n- Break= 
 11:15 – 11:45 am (chapter will provide pizza and soda)\n- Dr. Arindam Sa
 nyal (Arizona State University) lecture= 11:45 – 1:00pm\n\nBldg: Silicon
  Labs\, 200 W. Cezar Chavez\, Austin\, Texas\, United States\, 78701
LOCATION:Bldg: Silicon Labs\, 200 W. Cezar Chavez\, Austin\, Texas\, United
  States\, 78701
ORGANIZER:stefano.pietri@nxp.com
SEQUENCE:25
SUMMARY:Feed Your Mind - Practical Aspects of Machine Learning Circuits and
  Systems
URL;VALUE=URI:https://events.vtools.ieee.org/m/417876
X-ALT-DESC:Description: &lt;br /&gt;&lt;p class=&quot;Normal-FirstLine025&quot; style=&quot;text-in
 dent: 0in\;&quot;&gt;&lt;span style=&quot;font-family: Times New Roman\, serif\;&quot;&gt;&lt;span st
 yle=&quot;font-size: 14.6667px\;&quot;&gt;&lt;strong&gt;The &lt;strong style=&quot;mso-bidi-font-weig
 ht: normal\;&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; mso-bidi-font-size: 10.0pt\
 ; line-height: 105%\; font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;IEEE Centra
 l Texas CASS &amp;amp\; SSCS JT. Chapter\,&lt;/span&gt;&lt;/strong&gt;&lt;/strong&gt;&lt;/span&gt;&lt;/sp
 an&gt;&lt;/p&gt;\n&lt;p class=&quot;Normal-FirstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&lt;strong
  style=&quot;mso-bidi-font-weight: normal\;&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; m
 so-bidi-font-size: 10.0pt\; line-height: 105%\; font-family: &#39;Times New Ro
 man&#39;\,serif\;&quot;&gt;Circuits And Systems Society Outreach Program\,&lt;/span&gt;&lt;/str
 ong&gt;&lt;/p&gt;\n&lt;p class=&quot;Normal-FirstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&lt;stron
 g style=&quot;mso-bidi-font-weight: normal\;&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; 
 mso-bidi-font-size: 10.0pt\; line-height: 105%\; font-family: &#39;Times New R
 oman&#39;\,serif\;&quot;&gt;Silicon Laboratories &lt;/span&gt;&lt;/strong&gt;&lt;/p&gt;\n&lt;p class=&quot;Norma
 l-FirstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&lt;strong style=&quot;mso-bidi-font-we
 ight: normal\;&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; mso-bidi-font-size: 10.0p
 t\; line-height: 105%\; font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;Present:&lt;
 /span&gt;&lt;/strong&gt;&lt;/p&gt;\n&lt;p class=&quot;Normal-FirstLine025&quot; style=&quot;text-indent: 0i
 n\;&quot;&gt;Feed Your Mind - Practical Aspects of Machine Learning Circuits and S
 ystems&lt;/p&gt;\n&lt;p class=&quot;Normal-FirstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&lt;spa
 n style=&quot;font-size: 11.0pt\; mso-bidi-font-size: 10.0pt\; line-height: 105
 %\; font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;What is behind the buzzwords 
 Machine Learning\, Artificial Intelligence? In this lecture Prof. H. Li an
 d A. Sanyal will present practical aspects applied to circuits and systems
  solutions for everyday&#39;s life.&lt;/span&gt;&lt;/p&gt;\n&lt;p class=&quot;Normal-FirstLine025&quot;
  style=&quot;text-indent: 0in\;&quot;&gt;&lt;strong style=&quot;mso-bidi-font-weight: normal\;&quot;
 &gt;&lt;span style=&quot;font-size: 11.0pt\; mso-bidi-font-size: 10.0pt\; line-height
 : 105%\; font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;AI Models for Edge Compu
 ting: Hardware-aware Optimizations for Efficiency&lt;/span&gt;&lt;/strong&gt;&lt;/p&gt;\n&lt;p 
 class=&quot;Normal-FirstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&lt;span style=&quot;font-f
 amily: &#39;Times New Roman&#39;\,serif\;&quot;&gt;&amp;nbsp\;&lt;/span&gt;&lt;strong style=&quot;mso-bidi-f
 ont-weight: normal\;&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; mso-bidi-font-size:
  10.0pt\; line-height: 105%\; font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;&lt;em
 &gt;Abstract&lt;/em&gt;: &lt;span style=&quot;mso-spacerun: yes\;&quot;&gt;&amp;nbsp\;&lt;/span&gt;&lt;/span&gt;&lt;/s
 trong&gt;&lt;span style=&quot;font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;&amp;nbsp\;&lt;/span&gt;
 &lt;span style=&quot;font-family: &#39;Times New Roman&#39;\,serif\;&quot;&gt;As artificial intell
 igence (AI) transforms various industries\, state-of-the-art models have e
 xploded in size and capability. The growth in AI model complexity is rapid
 ly outstripping hardware evolution\, making the deployment of these models
  on edge devices remain challenging. To enable advanced AI locally\, model
 s must be optimized for fitting into the hardware constraints. In this pre
 sentation\, we will first discuss how computing hardware designs impact th
 e effectiveness of commonly used AI model optimizations for efficiency\, i
 ncluding techniques like quantization and pruning. Additionally\, we will 
 present several methods\, such as hardware-aware quantization and structur
 ed pruning\, to demonstrate the significance of software/hardware co-desig
 n. We will also demonstrate how these methods can be understood via a stra
 ightforward theoretical framework\, facilitating their seamless integratio
 n in practical applications and their straightforward extension to distrib
 uted edge computing. At the conclusion of our presentation\, we will share
  our insights and vision for achieving efficient and robust AI at the edge
 .&lt;/span&gt;&lt;/p&gt;\n&lt;p class=&quot;MsoNormal&quot;&gt;&lt;strong&gt;&lt;span style=&quot;font-size: 11.0pt\
 ; font-family: &#39;Times&#39;\,serif\; mso-bidi-font-family: &#39;Times New Roman&#39;\;&quot;
 &gt;Health management using intelligent wearables with mixed-signal AI&lt;/span&gt;
 &lt;/strong&gt;&lt;/p&gt;\n&lt;p class=&quot;MsoNormal&quot;&gt;&lt;span style=&quot;font-size: 11.0pt\; font-
 family: &#39;Times&#39;\,serif\; mso-bidi-font-family: &#39;Times New Roman&#39;\;&quot;&gt;&amp;nbsp\
 ;&lt;/span&gt;&lt;strong&gt;&lt;em&gt;&lt;span style=&quot;font-size: 11.0pt\; font-family: &#39;Times&#39;\
 ,serif\; mso-bidi-font-family: &#39;Times New Roman&#39;\;&quot;&gt;Abstract:&lt;/span&gt;&lt;/em&gt;&lt;
 /strong&gt;&lt;span style=&quot;font-size: 11.0pt\; font-family: &#39;Times&#39;\,serif\; mso
 -bidi-font-family: &#39;Times New Roman&#39;\;&quot;&gt; As medical wearables become more 
 widely adopted for at-home/early diagnosis/health surveillance\, the volum
 e of data produced by these devices are expected to reach thousands of pet
 abytes/month. Transmitting this large volume of data over the cloud for pr
 ocessing will potentially emerge as a communication bottleneck and increas
 e latency of decisions. Transmitting naively all data generated by a weara
 ble medical device is also costly in terms of power/energy- transmitter is
  usually the highest consumer of energy in a sensor (at least 10~20x more 
 energy than sensing). Key to addressing this data deluge is to increase ca
 pabilities of the wearable devices to process information locally and have
  on-device inference capabilities\, such as through embedding AI capabilit
 ies into the wearable device that will allow extraction of key information
  from the sensor data. There needs to be balance between what can be proce
 ssed locally on-device with low power/energy and how to optimally decide t
 he volume of data communication from the device (to cloud as an example). 
 The barriers to this approach lie in the computational complexity of AI al
 gorithms that makes it challenging to fit AI models on wearables with limi
 ted resources. Some of the answers might lie in going back to early days o
 f signal processing in silicon &amp;ndash\; developing analog circuit techniqu
 es for AI development which will require collaborative innovations in both
  AI model development and analog circuit design techniques. In this talk\,
  I will present our research on developing analog AI circuits and their de
 monstrations with patient data with use cases from cardiovascular health m
 onitoring and sepsis onset detection.&lt;/span&gt;&lt;/p&gt;&lt;br /&gt;&lt;br /&gt;Agenda: &lt;br /&gt;
 &lt;p class=&quot;xmsonormal&quot;&gt;Program:&lt;/p&gt;\n&lt;ul style=&quot;margin-top: 0in\;&quot; type=&quot;di
 sc&quot;&gt;\n&lt;li class=&quot;xmsolistparagraph&quot; style=&quot;margin-left: 0in\; mso-list: l0
  level1 lfo1\; tab-stops: list .5in\;&quot;&gt;&lt;span style=&quot;mso-fareast-font-famil
 y: &#39;Times New Roman&#39;\;&quot;&gt;Introduction Dr. S. Pietri (NXP) and J. Elenes (Si
 labs)=&amp;nbsp\; 10 &amp;ndash\; 10:15 am&lt;/span&gt;&lt;/li&gt;\n&lt;li class=&quot;xmsolistparagra
 ph&quot; style=&quot;margin-left: 0in\; mso-list: l0 level1 lfo1\; tab-stops: list .
 5in\;&quot;&gt;&lt;span style=&quot;mso-fareast-font-family: &#39;Times New Roman&#39;\;&quot;&gt;Dr. Hai 
 &quot;Helen&quot; Li (Duke University) lecture= 10-11:15 am&lt;/span&gt;&lt;/li&gt;\n&lt;li class=&quot;
 xmsolistparagraph&quot; style=&quot;margin-left: 0in\; mso-list: l0 level1 lfo1\; ta
 b-stops: list .5in\;&quot;&gt;&lt;span style=&quot;mso-fareast-font-family: &#39;Times New Rom
 an&#39;\;&quot;&gt;Break= 11:15 &amp;ndash\; 11:45 am (chapter will provide pizza and soda
 )&lt;/span&gt;&lt;/li&gt;\n&lt;li class=&quot;xmsolistparagraph&quot; style=&quot;margin-left: 0in\; mso
 -list: l0 level1 lfo1\; tab-stops: list .5in\;&quot;&gt;&lt;span style=&quot;mso-fareast-f
 ont-family: &#39;Times New Roman&#39;\;&quot;&gt;Dr. Arindam Sanyal (Arizona State Univers
 ity) lecture= 11:45 &amp;ndash\; 1:00pm&lt;/span&gt;&lt;/li&gt;\n&lt;/ul&gt;\n&lt;p class=&quot;Normal-F
 irstLine025&quot; style=&quot;text-indent: 0in\;&quot;&gt;&amp;nbsp\;&lt;/p&gt;
END:VEVENT
END:VCALENDAR