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:20260402T024534Z
LOCATION:3006\, Level 3
DTSTART;TZID=America/Los_Angeles:20250623T163000
DTEND;TZID=America/Los_Angeles:20250623T164500
UID:dac_DAC 2025_sess135_RESEARCH1217@linklings.com
SUMMARY:Real-Time Dynamic IR-drop Prediction for IR ECO
DESCRIPTION:Yu-Che Lee, Yu-Hsuan Chen, and Yu-Chen Cheng (National Tsing H
 ua University); Yong-Fong Chang, Jia Wei Lin, and Hsun-Wei Pao (MediaTek);
  Yung-Chih Chen (National Taiwan University of Science and Technology); an
 d Yi-Ting Li, Wuqian Tang, Shih-Chieh Chang, and Chun-Yao Wang (National T
 sing Hua University)\n\nDuring the IR Engineering Change Order (ECO) stage
 , cell moving leads to uncertain IR-drop results, requiring designers to e
 xplore multiple ECO candidates in each iteration to find a solution that e
 ffectively mitigates IR-drop, resulting in long evaluation time. Although 
 machine learning (ML)-based predictors have been proposed to expedite IR-d
 rop evaluation, partial simulations are still needed to update features af
 ter ECO, taking over an hour and delaying IR-drop results. In this work, w
 e propose a real-time dynamic IR-drop estimation method based on an XGBoos
 t model with a global view of a cell's surroundings. After ECO, our method
  provides dynamic IR-drop results in minutes without running any simulatio
 ns and thus achieve real-time estimation. This allows designers to evaluat
 e multiple ECO candidates concurrently in a single iteration. We conducted
  the experiments on five ECO candidates of an industrial design with 3𝑛𝑚 t
 echnology. The results show that the proposed model can effectively predic
 t the IR-drop variations of moved cells after ECO with over 93% of fixed c
 ells detected and an average MAE of 8.75𝑚𝑉 achieved. Furthermore, our meth
 od achieves an 88𝑋 speedup over Voltus (commercial tool) and a 64𝑋 speedup
  over traditional ML predictors when evaluating a single ECO candidate. Th
 e speedup is expected to increase as the number of ECO candidates increase
 s.\n\nTopics: EDA\n\nTracks: EDA4: Power Analysis and Optimization\n\nSess
 ion Chairs: Prabal Basu (Cadence Design Systems, Inc.) and Noel Daniel Gun
 di (Utah State University)\n\n
END:VEVENT
END:VCALENDAR