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:20181221T160728Z
LOCATION:D173
DTSTART;TZID=America/Chicago:20181112T140000
DTEND;TZID=America/Chicago:20181112T142000
UID:submissions.supercomputing.org_SC18_sess175_pec423@linklings.com
SUMMARY:HPC PowerStack: a community-wide open collaboration for enabling s
 ystem-wide power efficiency
DESCRIPTION:Workshop\nArchitectures, Collaborative Environments, Parallel 
 Programming Languages, Libraries, and Models, Simulation, Workflows, Works
 hop Reg Pass\n\nHPC PowerStack: a community-wide open collaboration for en
 abling system-wide power efficiency\n\nJana\n\nOne of the challenges to Op
 en Supercomputing in the exascale era is the design of software solutions 
 that drive Power Management across the system stack. In order to tackle th
 e power challenges, the HPC community has designed various open source and
  vendor-specific solutions to meet power management goals at various level
 s of granularity - from system level to node-level. These components are b
 uilt under certain assumptions of the system behavior and remain completel
 y agnostic of each other’s techniques. For example, an open source workloa
 d manager attempting to reduce the system-level power consumption may rema
 in completely oblivious of the power demands by the application-level runt
 ime and node-level hardware components. Attempting to integrate them into 
 a unified system stack often lead to them overstepping on each other’s sol
 utions which, in turn leads to unreliable system performance. To avoid thi
 s, system integrators end up designing their stack with tightly-coupled ch
 erry-picked vendor-specific solutions. This lack of coordination between t
 he vendors and the open solution community leads to underutilization of sy
 stem Watts and FLOPS. As a result, there is an urgent need for the HPC com
 munity to (A) identify the key software actors needed in a system power st
 ack: job-schedulers, application-level runtime, hardware knobs (B) arrive 
 at a consensus on the roles and responsibilities of these actors, (C) desi
 gn communication protocols for bidirectional control and feedback signals 
 among the actors to enable scalable closed-loop coordination at different 
 granularities, and (E) study and combine existing standalone engineering a
 nd development prototypes and build a community that actively participates
  in open development and engineering efforts.\n\nThis realization led to 
 the formation of the PowerStack Community, in 2016. The participants of th
 is community include global-wide members from academia, government labs, a
 nd vendors working on different layers of the system software stack. This 
 talk is intended to spread awareness among the workshop attendees and soli
 cit participation. The hope is to facilitate integrating open power manage
 ment solutions into future PowerStack-compatible systems.
URL:https://sc18.supercomputing.org/presentation/?id=pec423&sess=sess175
END:VEVENT
END:VCALENDAR