BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000611Z
LOCATION:704-706
DTSTART;TZID=America/Denver:20231112T140000
DTEND;TZID=America/Denver:20231112T173000
UID:submissions.supercomputing.org_SC23_sess437@linklings.com
SUMMARY:Workshop on Software and Hardware Co-Design of Deep Learning Syste
 ms in Accelerators (SHDA)
DESCRIPTION:SHDA – Afternoon Break\n---------------------\nWorkshop SHDA23
  Wrap-Up\n\nTong Shu (University of North Texas)\n---------------------\nP
 areto Optimization of CNN Models via Hardware-Aware Neural Architecture Se
 arch for Drainage Crossing Classification on Resource-Limited Devices\n\nE
 mbedded devices, constrained by limited memory and processors, require dee
 p learning models to be tailored to their specifications. This research ex
 plores customized model architectures for classifying drainage crossing im
 ages. Building on the foundational ResNet-18, this paper aims to maximize 
 pre...\n\n\nYuke Li (University of California, Merced); Jiwon Baik (Univer
 sity of California, Santa Barbara); Md Marufi Rahman (University of North 
 Texas); Iraklis Anagnostopoulos and Ruopu Li (Southern Illinois University
 ); and Tong Shu (University of North Texas)\n---------------------\nKeynot
 e: Design of Efficient and Privacy Preserving Machine Learning\n\nThe rapi
 d deployment of machine learning system has witnessed various challenges s
 uch as high computation and privacy/security concerns. In this talk, we wi
 ll first discuss the current challenges and advances in efficient machine 
 learning. We will present several machine learning accelerations throu...\
 n\n\nCaiwen Ding (University of Connecticut)\n---------------------\nBench
 marking and In-Depth Performance Study of Large Language Models on Habana 
 Gaudi Processors\n\nTransformer models suffer from high computational comp
 lexity. Habana GAUDI architecture offers a promising solution to tackle th
 ese issues. GAUDI features a Matrix Multiplication Engine (MME) and a clus
 ter of fully programmable Tensor Processing Cores (TPC). This paper explor
 es the untapped potentia...\n\n\nChengming Zhang and Baixi Sun (Indiana Un
 iversity); Xiaodong Yu (Stevens Institute of Technology, Argonne National 
 Laboratory (ANL)); Zhen Xie, Weijian Zheng, Kamil A. Iskra, and Pete Beckm
 an (Argonne National Laboratory (ANL)); and Dingwen Tao (Indiana Universit
 y)\n---------------------\nInvited Talk:  When Optimizing Software Produce
 s Optimized Hardware – A Case for Statically-Interpretable Control-Flow Pr
 ograms\n\nNowadays, powerful optimizing compilers are needed to transform 
 and specialize software for a particular machine, for performance and ener
 gy considerations. For example, compilers for High-level synthesis (HLS) c
 an greatly facilitate the description of complex hardware implementations,
  by raising th...\n\n\nLouis-Noel Pouchet (Colorado State University)\n---
 ------------------\nAccuracy-Constrained Efficiency Optimization and GPU P
 rofiling of CNN Inference for Detecting Drainage Crossing Locations\n\nThe
  accurate and efficient determination of hydrologic connectivity has garne
 red significant attention from both academic and industrial sectors due to
  its critical implications for environmental management. While recent stud
 ies have leveraged the spatial characteristics of hydrologic features, the
  ...\n\n\nYicheng Zhang (University of California, Riverside); Dhroov Pand
 ey (University of North Texas); Di Wu (Southern Illinois University); Turj
 a Kundu (University of North Texas); Ruopu Li (Southern Illinois Universit
 y); and Tong Shu (University of North Texas)\n---------------------\nWelco
 me to SC Workshop SHDA 2023\n\nTong Shu (University of North Texas)\n-----
 ----------------\nInvited Talk: I/O Profiling and Benchmarking for AI Appl
 ications\n\nTraining artificial intelligence (AI) models involves repeated
 ly loading large amounts of datasets. Data loading and transferring can po
 tentially become one of the bottlenecks. AI training has different Input/O
 utput (I/O) patterns compared with traditional scientific simulations. It 
 is read intensiv...\n\n\nHuihuo Zheng (Argonne National Laboratory (ANL))\
 n---------------------\nAccelerating Hyperparameter Optimization Algorithm
 s with Mixed Precision\n\nHyperparameter Optimization (HPO) of Neural Netw
 orks is a computationally expensive procedure, that has the potential to b
 enefit from the use of novel accelerator capabilities. This paper investig
 ates the performance of three popular HPO algorithms in terms of the achie
 ved speed-up and model accurac...\n\n\nMarcel Aach (Forschungszentrum Jüli
 ch, University of Iceland); Rakesh Sarma and Eray Inanc (Forschungszentrum
  Jülich); Morris Riedel (University of Iceland, Juelich Supercomputing Cen
 tre (JSC)); and Andreas Lintermann (Forschungszentrum Jülich)\n\nTag: Acce
 lerators, Codesign, Heterogeneous Computing, Task Parallelism\n\nRegistrat
 ion Category: Workshop Reg Pass\n\nSession Chairs: Tong Shu (University of
  North Texas), Seung-Hwan Lim (Oak Ridge National Laboratory (ORNL)), Pava
 n Balaji (Meta AI), and Sanjukta Bhowmick (University of North Texas)
END:VEVENT
END:VCALENDAR