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:20220812T074357Z
LOCATION:Osaka Room
DTSTART;TZID=Europe/Stockholm:20220629T110000
DTEND;TZID=Europe/Stockholm:20220629T130000
UID:submissions.pasc-conference.org_PASC22_sess167@linklings.com
SUMMARY:MS5A - Portable Solutions for High Energy Physics Workflows on Het
 erogeneous Architectures
DESCRIPTION:Minisymposium\n\nHigh-energy physics (HEP) experiments have de
 veloped millions of lines of code over decades that are optimized to run o
 n traditional x86 CPU systems. However we are seeing a rapidly increasing 
 fraction of floating point computing power in leadership-class computing f
 acilities and traditional data centers coming from new heterogeneous accel
 erator architectures, such as GPUs. Though the GPU field is currently bein
 g led by NVIDIA, other manufacturers such as Intel and AMD are making incr
 easing inroads into this territory, each with their own architecture and c
 ompiler languages. Many HEP experiments are also using FPGAs for their fro
 ntend detector readouts. Rewriting current CPU based High Energy Physics c
 ode for multiple accelerator architectures is not a viable scenario, given
  the available person power and code maintenance issues. Furthermore, as t
 he number of architectures proliferate, it becomes increasingly onerous to
  validate the code, and it is vital to ensure that workflows on different 
 hardware produce identical results. Developing portable solutions for HEP 
 codebases to run on multiple heterogeneous architectures is essential. We 
 explore how major HEP experiments are addressing these issues, and also a 
 domain wide investigation to compare the efficacy of the various portabili
 ty layers such as Kokkos, SYCL, Alpaka, OpenMP, and std::execution::parall
 el.\n\nVecMem: A Library for Managing Heterogeneous Application Memory Usi
 ng Modern C++\n\nKrasznahorkay, Swatman, Gessinger\n\nUsing modern, high-l
 evel C++ code in heterogeneous programming is made difficult by the extra 
 dimensions of phase space used in managing the memory of the code. In orde
 r to use the memory of non-CPU devices, one generally needs to make use of
  allocation/deallocation functions fitting a C application...\n\n---------
 ------------\nA Heterogeneous Trigger System for Upgrade LHCb\n\nGligorov\
 n\nIn 2022 the upgraded LHCb experiment will use a triggerless readout sys
 tem collecting data at an event rate of 30 MHz and a data rate of 4 Teraby
 tes/second. A software-only High Level Trigger will enable unprecedented f
 lexibility for trigger selections. During the first stage (HLT1), track re
 constru...\n\n---------------------\nDeveloping Parallelization Strategies
  and Evaluating Portability Layers for High Energy Physics Experiments\n\n
 Lin, Leggett\n\nHEP is faced with the untenable prospect of rewriting mill
 ions of lines of code, which have been optimized for x86 CPUs, for the inc
 reasingly dominant architectures found in computational accelerators such 
 as GPUs. This task is made worse by the architecture specific languages an
 d APIs promoted by m...\n\n---------------------\nFPGAs for ML Inference i
 n CMS\n\nJames\n\nThe Level-1 (L1) Trigger of the CMS experiment at the CE
 RN LHC selects potentially interesting particle collision events at a rate
  of approximately 40 MHz, within a timespan of a few microseconds. As a re
 sult of the huge bandwidth requirements, and tiny latency budget, this ste
 p of the online proces...\n\n\nDomain: Computer Science and Applied Mathem
 atics, Physics
END:VEVENT
END:VCALENDAR