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:20260625T133341Z
LOCATION:Bldg. 6 - 001 - Plenary Room
DTSTART;TZID=Europe/Stockholm:20260630T124200
DTEND;TZID=Europe/Stockholm:20260630T124300
UID:submissions.pasc-conference.org_PASC26_sess129_posC104@linklings.com
SUMMARY:ACMP13 - Solver-Integrated Lossy and Lossless Compression for Scal
 able Flow Simulations
DESCRIPTION:Viral Sudip Shah (University of Illinois Urbana-Champaign)\n\n
 Large-scale computational fluid dynamics (CFD) simulations on modern GPU-a
 ccelerated supercomputers generate terabytes of data per run, making check
 pointing, storage, and post-hoc analysis increasingly I/O-bound. This bott
 leneck limits data retention and hinders downstream workflows such as visu
 alization and data-driven modeling. In this work, we integrate both lossle
 ss and error-bounded lossy compression directly into the GPU-enabled spect
 ral element solver nekRS, enabling scalable, solver-integrated data reduct
 ion with minimal communication overhead. Our approach adopts an embarrassi
 ngly parallel design in which each MPI rank independently compresses its l
 ocal solution fields during checkpointing and writes compressed buffers us
 ing MPI-IO. We evaluate lossless compression using Blosc2 for restart file
 s and lossy compression using SZ3 for analysis-oriented outputs. Results f
 rom a large-scale jet-in-crossflow simulation demonstrate that lossy compr
 ession achieves compression ratios exceeding 100x while preserving key flo
 w structures and vortex topology, and reduces I/O time by up to 8x compare
 d to uncompressed output. Lossless compression provides bitwise reproducib
 ility with high throughput for reliable restarts. These results show that 
 solver-integrated compression can significantly alleviate I/O bottlenecks 
 in large-scale CFD and enable more efficient data-centric workflows on eme
 rging HPC systems.\n\nSession Chair: Tobias Hodel (University of Bern, Swi
 tzerland)\n\n
END:VEVENT
END:VCALENDAR