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:20230124T171519Z
LOCATION:D222
DTSTART;TZID=America/Chicago:20221113T160000
DTEND;TZID=America/Chicago:20221113T163000
UID:submissions.supercomputing.org_SC22_sess436_ws_mchpc103@linklings.com
SUMMARY:Reducing Memory-Bus Energy Consumption of GPUs via Software-Based 
 Bit-Flip Minimization
DESCRIPTION:Workshop\n\nReducing Memory-Bus Energy Consumption of GPUs via
  Software-Based Bit-Flip Minimization\n\nFallin, Burtscher\n\nEnergy consu
 mption is a major concern in high-performance computing. One important con
 tributing factor is the number of times the wires are charged and discharg
 ed, i.e., how often they switch from '0' to '1' and vice versa. We describ
 e a software technique to minimize this switching activity in GPUs, thereb
 y lowering the energy usage. Our technique targets the memory bus, which c
 omprises many high-capacitance wires that are frequently used. Our approac
 h is to strategically change data values in the source code such that load
 ing and storing them yields fewer bit flips. The new values are guaranteed
  to produce the same control flow and program output. Measurements on GPUs
  from two generations show that our technique allows programmers to save u
 p to 9.3% of the whole-GPU energy consumption and 1.2% on average across e
 ight graph-analytics CUDA codes without impacting performance.\n\nSession 
 Format: Recorded\n\nRegistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR