BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211207T055407Z
LOCATION:Online
DTSTART;TZID=America/Chicago:20211115T090000
DTEND;TZID=America/Chicago:20211115T174000
UID:submissions.supercomputing.org_SC21_sess423@linklings.com
SUMMARY:7th Workshop on Machine Learning in High Performance Environment
DESCRIPTION:Workshop\n\nHigh-Performance Deep Learning Toolbox for Genome-
 Scale Prediction of Protein Structure and Function\n\nGao, Sedova, Cheng\n
 \nComputational biology is one of many scientific disciplines ripe for inn
 ovation and acceleration with the advent of high-performance computing (HP
 C). In recent years, the field of machine learning has also seen significa
 nt benefits from adopting HPC practices. In this work, we present a novel 
 HPC pi...\n\n---------------------\nMorning Invited Talk\n\nMesser\n\n----
 -----------------\nAfternoon Invited Talk\n\nCatanzaro\n\n----------------
 -----\nHPC Ontology: Toward a Unified Ontology for Managing Training Datas
 ets and AI Models for High-Performance Computing\n\nLiao, Lin, Verma, Vand
 erbruggen, Emani...\n\nMachine learning (ML) techniques have been widely s
 tudied to address various challenges of productively and efficiently runni
 ng large-scale scientific applications on heterogeneous supercomputers. Ho
 wever, it is extremely difficult to generate, access, and maintain trainin
 g datasets and AI models to...\n\n---------------------\nIs Disaggregation
  Possible for HPC Cognitive Simulation?\n\nWyatt, Yamamoto, Tosi, Karlin, 
 Van Essen\n\nCognitive simulation (CogSim) is an important and emerging wo
 rkflow for HPC scientific exploration and scientific machine learning (Sci
 ML).   One challenging workload for CogSim is the replacement of one compo
 nent in a complex physical simulation with a fast, learned, surrogate mode
 l that is inside ...\n\n---------------------\nHYPPO: A Surrogate-Based Mu
 lti-Level Parallelism Tool for Hyperparameter Optimization\n\nDumont\n\nWe
  present a new software, HYPPO, that enables the automatic tuning of hyper
 parameters of various deep learning models. Unlike other hyperparameter op
 timization methods, HYPPO uses adaptive surrogate models and directly acco
 unts for uncertainty in model predictions to find accurate and reliable mo
 de...\n\n---------------------\n7th Workshop on Machine Learning:  Afterno
 on Break (3-3:30)\n\n\n\n---------------------\nHPCFAIR: Enabling FAIR AI 
 for HPC Applications\n\nVerma, Emani, Liao, Lin, Vanderbruggen...\n\nArtif
 icial Intelligence (AI) is being adopted in different domains at an unprec
 edented scale. A significant interest in the scientific community also inv
 olves leveraging machine learning (ML) to run high-performance computing a
 pplications at scale effectively. Given multiple efforts in this arena, t.
 ..\n\n---------------------\nSemantic-Aware Lossless Data Compression for 
 Deep Learning Recommendation Model (DLRM)\n\nPumma, Vishnu\n\nDeep Learnin
 g Recommendation Model (DLRM), a new neural network for recommendation sys
 tems, introduces challenging requirements for deep neural network training
  and inference.  The size of the DLRM model is typically large and not abl
 e to fit on a single GPU memory.  DLRM requires both model-paralle...\n\n-
 --------------------\nColmena: Scalable Machine-Learning-Based Steering of
  Ensemble Simulations for High Performance Computing\n\nWard, Sivaraman, P
 auloski, Babuji, Chard...\n\nScientific applications that involve simulati
 on ensembles can be accelerated greatly by using experiment design methods
  to select the best simulations to perform. Methods that use machine learn
 ing (ML) to create proxy models of simulations show particular promise for
  guiding ensembles but are challe...\n\n---------------------\nProduction 
 Deployment of Machine-Learned Rotorcraft Surrogate Models on HPC\n\nBrewer
 \n\nWe explore how to optimally deploy different types of machine-learned 
 surrogate models used in rotorcraft aerodynamics on HPC. We first develope
 d three different rotorcraft models at three different orders of magnitude
  (2M, 44M, and 212M trainable parameters) to use as test models. We tested
  three d...\n\n---------------------\n7th Workshop on Machine Learning:  M
 orning Break (10-10:30)\n\n\n\n---------------------\nWelcome:  Workshop o
 n Machine Learning in High Performance Computing Environments\n\nLim\n\n--
 -------------------\n7th Workshop on Machine Learning in High Performance 
 Environment\n\nLim, Keuper, Houston, Patton\n\nThe intent of this workshop
  is to bring together researchers, practitioners and scientific communitie
 s to discuss methods that utilize extreme scale systems for machine learni
 ng. This workshop will focus on the greatest challenges in utilizing HPC f
 or machine learning and methods for exploiting data...\n\n----------------
 -----\n7th Workshop on Machine Learning:  Lunch Break (12:30-2)\n\n\n\n---
 ------------------\nMLPerf HPC: A Holistic Benchmark Suite for Scientific 
 Machine Learning on HPC Systems\n\nFarrell, Emani, Balma, Drescher, Drozd.
 ..\n\nScientific communities are increasingly adopting machine learning an
 d deep learning models in their applications to accelerate scientific insi
 ghts. High performance computing systems are pushing the frontiers of perf
 ormance with a rich diversity of hardware resources and massive scale-out 
 capabiliti...\n\n\nTag: Online Only, Machine Learning and Artificial Intel
 ligence\n\nRegistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR