BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250626T233526Z
LOCATION:B302-B305
DTSTART;TZID=America/New_York:20241120T100000
DTEND;TZID=America/New_York:20241120T170000
UID:submissions.supercomputing.org_SC24_sess533_post133@linklings.com
SUMMARY:An Adaptive Kernel Execution for Dynamic Applications on GPUs Usin
 g CUDA Graphs
DESCRIPTION:Kento Kitamura, Kenji Tanaka, and Kazunori Seno (NTT Corporati
 on)\n\nWe propose a novel approach for executing dynamic applications on G
 PUs. Different from traditional approaches that use a single kernel, our m
 ethod allows the GPU to autonomously allocate computational resources at r
 untime. We decompose a kernel into multiple fragment kernels and dynamical
 ly launch an optimal number of them during execution. The input data is pa
 rtitioned into smaller segments and each fragmented kernel processes each 
 of the partitioned segments. This method is implemented using CUDA graphs 
 conditional nodes for determining the number of fragmented kernels to be l
 aunched based on the input size. We compared the proposed method with the 
 traditional kernel execution method with a Breadth-First Search (BFS) appl
 ication, a representative dynamic application. Results show comparable per
 formance while reducing utilization of compute resources by up to 19.9%, a
 nd opportunities to further performance improvement by optimizing paramete
 rs of our method.\n\nRegistration Category: Tech Program Reg Pass, Exhibit
 s Reg Pass\n\nSession Chairs: Ayesha Afzal (Friedrich-Alexander University
 , Erlangen-Nuremberg; Erlangen National High Performance Computing Center)
 ; Sally Ellingson (University of Kentucky); and Alan Sussman (University o
 f Maryland)\n\n
END:VEVENT
END:VCALENDAR