BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260421T090514Z
LOCATION:Bldg. 6 - Room 102
DTSTART;TZID=Europe/Stockholm:20260629T172000
DTEND;TZID=Europe/Stockholm:20260629T180000
UID:submissions.pasc-conference.org_PASC26_sess119_msa165@linklings.com
SUMMARY:ProACT: Fair and Carbon-Aware Scheduling for Responsible HPC
DESCRIPTION:Denisa-Andreea Constantinescu (EPFL) and Steven Senator (Los A
 lamos National Laboratory)\n\nHPC infrastructures face increasing pressure
  to reduce emissions while continuing to support diverse and growing scien
 tific workloads. Initiatives by major cloud providers such as Google, Micr
 osoft, and Amazon demonstrate that delaying or relocating flexible workloa
 ds can significantly reduce emissions. Emerging platforms such as EmeraldA
 I further explore datacenter-level flexibility by adapting computing deman
 d to grid conditions. However, these approaches generally assume workload 
 flexibility rather than enabling it explicitly within scheduling systems. 
 At the same time, evidence suggests that many users are willing to adopt g
 reener computing practices—such as accepting alternative resources, schedu
 ling policies, or quality of service—when given the opportunity.\n\nThis r
 aises an important question for HPC schedulers: what capabilities are requ
 ired to allow users to express such preferences and participate in carbon-
 aware resource management? If users accept delayed execution to support ca
 rbon-aware scheduling, how should compute usage, emissions, and resource a
 llocations be transparently attributed?\n\nIn this talk, we examine what c
 apabilities HPC schedulers would need to enable user-declared flexibility 
 while maintaining fairness and transparency in resource allocation and emi
 ssions accounting. Specifically, we analyze different workload flexibility
  modes using insights from the SEAMS project (Sustainable & Energy Aware M
 ethods for the SKA Observatory) and emerging large-scale AI workloads.\n\n
 Domain: Climate, Weather, and Earth Sciences, Engineering, Life Sciences, 
 Computational Methods and Applied Mathematics\n\nSession Chairs: Maria Gra
 zia Giuffreda (ETH Zurich / CSCS), Florina Ciorba (University of Basel), M
 arie-Christine Sawley (ICES Foundation), and Maria Girone (CERN)\n\n
END:VEVENT
END:VCALENDAR