BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211208T164406Z
LOCATION:Level 2 - Exhibit Hall
DTSTART;TZID=America/Los_Angeles:20211206T170000
DTEND;TZID=America/Los_Angeles:20211206T180000
UID:dac_DAC 2021_sess249_DESPOST114@linklings.com
SUMMARY:Accelerating Standard Cells Variation-Aware Characterization Metho
 dology with Machine Learning Techniques
DESCRIPTION:Designer, IP and Embedded Systems Track Poster Networking Rece
 ption\n\nAccelerating Standard Cells Variation-Aware Characterization Meth
 odology with Machine Learning Techniques\n\nTan\n\nVariation-aware charact
 erization for timing libraries (.libs) is a necessity for modern process n
 odes, due to the increased impact of process variability from shrinking fe
 ature size, increased device integration and low voltage operations. \n\nA
  key hurdle to producing accurate variation model timing views is the runt
 ime required to characterize Liberty Variation Format (LVF) .libs. Due to 
 the millions of data points required over many process, voltage and temper
 ature (PVT) corners, a full characterization run may require weeks or even
  months to complete. \n\nHowever, the abundance of data points in .libs al
 so present an opportunity to leverage machine learning (ML) methods to pro
 duce accurate characterization results while significantly reducing total 
 characterization runtime. \n\nIn this paper, we present the usage and resu
 lts of machine learning methods that complement SPICE characterization of 
 LVF .libs to improve characterization runtime while maintaining equivalent
  output .lib accuracy. Results expected are on the order of ~100X speedup 
 for direct runtime comparison between SPICE characterization and ML genera
 tion per .lib, and 40%-60% runtime reduction across a full set of .libs, d
 epending on number of .libs generated using ML. We will present actual res
 ults from our testcase in this paper.\n\nVirtual Program: Presented In-Per
 son
END:VEVENT
END:VCALENDAR