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:20250626T234541Z
LOCATION:B308
DTSTART;TZID=America/New_York:20241119T103000
DTEND;TZID=America/New_York:20241119T110000
UID:submissions.supercomputing.org_SC24_sess377_pap223@linklings.com
SUMMARY:High-ratio Scientific Lossy Compression on GPUs with Optimized Mul
 ti-level Interpolation
DESCRIPTION:Jinyang Liu (University of California, Riverside); Jiannan Tia
 n (Indiana University); Shixun Wu (University of California, Riverside); S
 heng Di (Argonne National Laboratory (ANL)); Boyuan Zhang (Indiana Univers
 ity); Robert Underwood (Argonne National Laboratory (ANL)); Yafan Huang (T
 he University of Iowa); Jiajun Huang (University of California, Riverside)
 ; Kai Zhao (Florida State University); Guanpeng Li (The University of Iowa
 ); Dingwen Tao (Indiana University); Zizhong Chen (University of Californi
 a, Riverside); and Franck Cappello (Argonne National Laboratory (ANL))\n\n
 Error-bounded lossy compression is a critical technique for significantly 
 reducing scientific data volumes. Compared to CPU-based compressors, GPU-b
 ased compressors exhibit substantially higher throughputs, fitting better 
 for today's HPC applications. However, the critical limitations of existin
 g GPU-based compressors are their low compression ratios and qualities, se
 verely restricting their applicability. To overcome these, we introduce a 
 novel GPU-based error-bounded scientific lossy compressor named cuSZ-I, wi
 th the following contributions: (1) A novel GPU-optimized interpolation-ba
 sed prediction method significantly improves the compression ratio and dec
 ompression data quality. (2) The Huffman encoding module in cuSZ-I is opti
 mized for better efficiency. (3) cuSZ-I is the first to integrate the NVID
 IA Bitcomp-lossless as an additional compression-ratio-enhancing module. E
 valuations show that cuSZ-I significantly outperforms other latest GPU-bas
 ed lossy compressors in compression ratio under the same error bound (henc
 e, the desired quality), showcasing a 476% advantage over the second-best.
  This leads to cuSZ-I's optimized performance in several real-world use ca
 ses.\n\nTag: Accelerators, Algorithms, Data Compression, I/O, Storage, Arc
 hive, Performance Optimization\n\nRegistration Category: Tech Program Reg 
 Pass\n\nSession Chair: Ganesh Gopalakrishnan (University of Utah)\n\n
END:VEVENT
END:VCALENDAR