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:20260202T201804Z
LOCATION:261
DTSTART;TZID=America/Chicago:20251117T155500
DTEND;TZID=America/Chicago:20251117T162500
UID:submissions.supercomputing.org_SC25_sess214_ws_corr103@linklings.com
SUMMARY:Towards an Automated Workflow for Floating-Point Analysis of GPU K
 ernels
DESCRIPTION:Esteban M. Rangel (Argonne National Laboratory (ANL)) and S. J
 ohn Pennycook (Intel Corporation)\n\nWe present early work towards establi
 shing an automated workflow for numerical error analysis of SYCL kernels. 
 The method leverages the OpenCL Intercept Layer to record the execution of
  GPU-targeted SYCL kernels and replay them with CPU runtimes, enabling det
 ailed floating-point error evaluation without modifying the original appli
 cation. We analyze the force kernel from a large-scale cosmology applicati
 on code (HACC) using PoCL and Verificarlo, exploring both IEEE-compliant c
 onfigurations and reduced-precision modes (\eg, FP16, TF32, BF16), as well
  as the effects of the \texttt{-ffast-math} compiler optimization using st
 ochastic arithmetic via MCA and PRISM. By leveraging open tools and standa
 rds, this work contributes a reusable path toward broader adoption of nume
 rical accuracy evaluation in HPC.\n\nTag: Debugging & Correctness Tools, H
 PC Software & Runtime Systems\n\nRecording: Livestreamed, Recorded\n\nRegi
 stration Category: Technical Program Reg Pass, Workshop Reg Pass\n\nSessio
 n Chairs: Ignacio Laguna (Lawrence Livermore National Laboratory (LLNL)) a
 nd Cindy Rubio-González (University of California, Davis)\n\n
END:VEVENT
END:VCALENDAR