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:20241120T082409Z
LOCATION:HG E 1.2
DTSTART;TZID=Europe/Stockholm:20240605T120000
DTEND;TZID=Europe/Stockholm:20240605T123000
UID:submissions.pasc-conference.org_PASC24_sess127_msa154@linklings.com
SUMMARY:Making Fortran Fly on AMD Instinct Accelerators
DESCRIPTION:Minisymposium\n\nMichael Klemm (AMD, OpenMP) and Paul Mullowne
 y (AMD)\n\nFortran has played and is playing a vital tole in the HPC ecosy
 stem, especially in the domain of weather forecasting.  AMD is a key contr
 ibutor to the LLVM/Flang compiler infrastructure and plays a key role in e
 stablishing an LLVM-based compiler for Fortran for both CPU and GPU.  In t
 his talk, we will briefly recap the AMD EPYC(tm) Processor architecture an
 d AMD Instinct(tm) GPU and APU platform as well as the available AMD softw
 are ecosystem. We will then turn towards how to utilize AMD GPUs and APUs 
 from Fortran code and how to offload compute kernels to GPUs and APUs. In 
 particular, we will focus on the ECMWF CloudSC microbenchmarks, i.e., a se
 t of resource (register) hungry compute kernels, as our barometer of offlo
 aded kernel compute performance. We will show some of the tuning opportuni
 ties available to programmers and compare performance with native HIP impl
 ementation on MI250X and MI300A architectures.\n\nDomain: Climate, Weather
 , and Earth Sciences, Computational Methods and Applied Mathematics\n\nSes
 sion Chairs: Michael Lange (ECMWF) and Balthasar Reuter (ECMWF)
END:VEVENT
END:VCALENDAR