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:20260522T150123Z
LOCATION:D167/174
DTSTART;TZID=America/Chicago:20181111T160000
DTEND;TZID=America/Chicago:20181111T163000
UID:submissions.supercomputing.org_SC18_sess221_ws_mlhpce103@linklings.com
SUMMARY:Auto-Tuning TensorFlow Threading Model for CPU Backend
DESCRIPTION:Niranjan Hasabnis (Intel Corporation)\n\nTensorFlow is a popul
 ar deep learning framework used to solve machine learning and deep learnin
 g problems such as image classification and speech recognition. It also al
 lows users to train neural network models or deploy them for inference usi
 ng GPUs, CPUs, and custom-designed hardware such as TPUs. Even though Tens
 orFlow supports a variety of optimized backends, realizing the best perfor
 mance using a backend requires additional efforts. Getting the best perfor
 mance from a CPU backend requires tuning of its threading model. Unfortuna
 tely, the best tuning approach used today is manual, tedious, time-consumi
 ng, and, more importantly, may not guarantee the best performance.\n\nIn t
 his paper, we develop an automatic approach, called TENSORTUNER, to search
  for optimal parameter settings of TensorFlow’s threading model for CPU ba
 ckends. We evaluate TENSORTUNER on both Eigen and Intel’s MKL CPU backends
  using a set of neural networks from TensorFlow’s benchmarking suite. Our 
 evaluation results demonstrate that the parameter settings found by TENSOR
 TUNER produce 2% to 123% performance improvement for the Eigen CPU backend
  and 1.5% to 28% performance improvement for the MKL CPU backend over the 
 performance obtained using their best-known parameter settings. This highl
 ights the fact that the default parameter settings in Eigen CPU backend ar
 e not the ideal settings; and even for a carefully hand-tuned MKL backend,
  the settings are sub-optimal. Our evaluations also revealed that TENSORTU
 NER is efficient at finding the optimal settings — it is able to converge 
 to the optimal settings quickly by pruning more than 90% of the parameter 
 search space.\n\nTag: Applications, Deep Learning, Machine Learning\n\nReg
 istration Category: Workshop Reg Pass\n\n
END:VEVENT
END:VCALENDAR