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:20260202T201337Z
LOCATION:265
DTSTART;TZID=America/Chicago:20251117T090000
DTEND;TZID=America/Chicago:20251117T173000
UID:submissions.supercomputing.org_SC25_sess217@linklings.com
SUMMARY:The 11th International Workshop on Data Analysis and Reduction for
  Big Scientific Data
DESCRIPTION:In this new exascale computing era, applications must increasi
 ngly perform online data analysis and reduction — tasks that introduce alg
 orithmic, implementation, and programming model challenges unfamiliar to m
 any scientists and with major implications for the design and use of vario
 us elements of exascale systems. There are at least three important topics
  that this workshop is striving to address: (1) whether several orders of 
 magnitude of data reduction is possible for exascale sciences; (2) underst
 anding the performance and accuracy trade-off of data reduction; and (3) s
 olutions to effectively reduce data while preserving the information hidde
 n in large scientific datasets. Tackling these challenges requires experti
 se from computer science, mathematics, and application domains to study th
 e problem holistically and develop solutions and robust software tools.\n\
 nThe 11th International Workshop on Data Analysis and Reduction for Big Sc
 ientific Data\n\nIn this new exascale computing era, applications must inc
 reasingly perform online data analysis and reduction — tasks that introduc
 e algorithmic, implementation, and programming model challenges unfamiliar
  to many scientists and with major implications for the design and use of 
 various element...\n\n\nSheng Di (Argonne National Laboratory (ANL)), Ana 
 Gainaru (Oak Ridge National Laboratory (ORNL)), Kento Sato (RIKEN Center f
 or Computational Science (R-CCS)), and Xin Liang (University of Kentucky)\
 n---------------------\nLightweight CNN-Based Artifact Reduction for Scien
 tific Error-bounded Lossy Compression\n\nLossy compression is widely used 
 to reduce storage and transmission costs in large-scale scientific data, b
 ut it inevitably introduces artifacts that may compromise subsequent analy
 sis. To address this issue, we propose a lightweight 3D convolutional arch
 itecture with a fixed-scale batch normalizati...\n\n\nZizhe Jian (Universi
 ty of California, Riverside); Pu Jiao (University of Kentucky); Bohan Zhan
 g (University of Florida); Sheng Di (Argonne National Laboratory (ANL)); X
 in Liang (University of Kentucky); Guanpeng Li (University of Florida); Hu
 angliang Dai and Zizhong Chen (University of California, Riverside); and F
 ranck Cappello (Argonne National Laboratory (ANL))\n---------------------\
 nASCRIBE-XR: Extended Reality for Visualization of Scientific Images\n\nWe
  introduce ASCRIBE-XR, an immersive software application designed to accel
 erate the visualization and exploration of 3D dense arrays and mesh files 
 from scientific experiments. Based on Godot and PC-VR technologies, the pl
 atform enables users to dynamically load and manipulate scientific records
  t...\n\n\nRonald J. Pandolfi (Lawrence Berkeley National Laboratory (LBNL
 )); Julian Todd (DOESLiverpool); Jeffrey Donatelli (Lawrence Berkeley Nati
 onal Laboratory (LBNL)); and Daniela Ushizima (Lawrence Berkeley National 
 Laboratory (LBNL), University of California San Francisco)\n--------------
 -------\nEvaluating Accuracy and Performance Tradeoffs in GPU Accelerated 
 Single Cell RNA-seq Analysis\n\nSingle-cell RNA sequencing (scRNA-seq) now
  profiles millions of cells in a single study, creating major computationa
 l demands. GPU-accelerated pipelines, built on frameworks like NVIDIA RAPI
 DS and CuPy, promise large runtime reductions, but questions remain about 
 reproducibility compared to CPU work...\n\n\nCory Gardner, Seyun Jeong, an
 d Oam Khatavkar (Saint Louis University); Aiden Moon (Parkway Central High
  School); and Qinglei Cao and Tae-Hyuk Ahn (Saint Louis University)\n-----
 ----------------\nAfternoon Break - Workshop on Data Analysis and Reductio
 n for Big Scientific Data\n---------------------\nOn the Compressibility o
 f Floating-Point Data in Posit and IEEE-754 Representation\n\nThe IEEE 754
  floating-point standard is the most used representation for real numbers 
 in modern computer systems, despite issues in accuracy for certain applica
 tions. The posit format, which has several advantages, has been proposed a
 s a direct drop-in replacement for IEEE floats. Many works compare...\n\n\
 nAndrew Rodriguez and Martin Burtscher (Texas State University)\n---------
 ------------\nMorning Break - Workshop on Data Analysis and Reduction for 
 Big Scientific Data\n---------------------\nFZModules: A Heterogeneous Com
 puting Framework for Customizable Scientific Data Compression Pipelines\n\
 nModern scientific simulations and instruments generate data volumes that 
 overwhelm memory and storage, throttling scalability. Lossy compression mi
 tigates this by trading controlled error for reduced footprint and through
 put gains, yet optimal pipelines are highly data and objective specific, d
 emand...\n\n\nSkyler Ruiter (Indiana University), Jiannan Tian (Oakland Un
 iversity), and Fengguang Song (Indiana University)\n---------------------\
 nData Management System Analysis for Distributed Computing Workloads\n\nLa
 rge-scale international collaborations such as ATLAS rely on globally dist
 ributed workflows and data management to process, move, and store vast vol
 umes of data. ATLAS’s Production and Distributed Analysis (PanDA) workflow
  system and the Rucio data management system are each highly optimized...\
 n\n\nKuan-Chieh Hsu, Sairam Sri Vatsavai, Ozgur O. Kilic, Sankha Dutta, Yi
 hui (Ray) Ren, and David Park (Brookhaven National Laboratory); Tania Korc
 huganova and Joseph Boudreau (University of Pittsburgh); Tasnuva Chowdhury
  (Brookhaven National Laboratory); Shengyu Feng (Carnegie Mellon Universit
 y); Raees Ahmad Khan (University of Pittsburgh); Jaehyung Kim (Carnegie Me
 llon University); Norbert Podhorszki and Scott Klasky (Oak Ridge National 
 Laboratory (ORNL)); Tadashi Maeno, Paul Nilsson, and Verena Ingrid Martine
 z Outschoorn (University of Pittsburgh); Fred Suter (Oak Ridge National La
 boratory (ORNL)); Wei Yang (SLAC National Accelerator Laboratory); Yiming 
 Yang (Carnegie Mellon University); and Shinjae Yoo, Alexei Klimentov, and 
 Adolfy Hoisie (Brookhaven National Laboratory)\n---------------------\nCha
 racterizing the Performance of Parallel Data-Compression Algorithms across
  Compilers and GPUs\n\nDifferent compilers can generate code with notably 
 different performance characteristics—even on the same system. Today, GPU 
 developers have three popular options for compiling CUDA or HIP code for G
 PUs. First, CUDA code can be compiled by either NVCC or Clang for NVIDIA G
 PUs. Alternatively, A...\n\n\nBrandon Alexander Burtchell and Martin Burts
 cher (Texas State University)\n---------------------\nBenchmarking Cutting
 -Edge Scientific Error-Bounded Lossy Compressors on Correlation-Based Rate
 -Distortion\n\nScientific error-bounded lossy compressors are widely used 
 to reduce storage and I/O costs in large-scale scientific computing tasks.
  It is critical to benchmark those compressors to help users understand th
 eir performance. Nevertheless, when evaluating the decompressed data quali
 ty, existing benchm...\n\n\nZiwei Qiu (University of Houston)\n-----------
 ----------\nInvited Talk: Globus: Enabling Scalable and Sustainable Resear
 ch for Data-Intensive Science\n\nScientific research increasingly depends 
 on the movement, management, and analysis of massive data volumes. Globus,
  a widely used research IT platform, addresses these needs by providing se
 cure, reliable, and high-performance capabilities for data management, com
 putation, and workflows across global...\n\n\nKyle Chard (University of Ch
 icago, Argonne National Laboratory (ANL))\n---------------------\nCompress
 ion Error Sensitivity Analysis for Different Experts in MoE Model Inferenc
 e\n\nWith the widespread application of Mixture of Experts (MoE) reasoning
  models in the field of LLM learning, efficiently serving MoE models under
  limited GPU memory constraints has emerged as a significant challenge. Of
 floading the non-activated experts to main memory has been identified as a
 n efficie...\n\n\nSongkai Ma (Hong Kong Polytechnic University); Zhaorui Z
 hang (The Hong Kong Polytechnic University); Sheng Di (Argonne National La
 boratory (ANL)); Benben Liu (The University of Hong Kong); Xiaodong Yu (St
 evens Institute of Technology); Xiaoyi Lu (University of California, Merce
 d); and Dan Wang (Hong Kong Polytechnic University)\n---------------------
 \nBuilding n-Dimensional Trees for Resolution-Based Progressive Compressio
 n\n\nFloating-point data is typically compressed at strict error bounds to
  reduce storage cost while facilitating scientific analyses. Unfortunately
 , this tends to yield large compressed files. In some cases, however, a us
 er might not need the data at a high fidelity. Progressive compression add
 resses th...\n\n\nBrandon Alexander Burtchell and Martin Burtscher (Texas 
 State University)\n---------------------\nIntegrating Distributed SQL Quer
 y Engines with Object-Based Computational Storage\n\nExisting object stora
 ge systems like AWS S3 and MinIO offer only limited in-storage compute cap
 abilities, typically restricted to simple SQL WHERE-clause filtering. Cons
 e-\nquently, high-impact operators—such as aggregation and top-N—are still
  executed entirely at the compute layer. Recen...\n\n\nJunghyun Ryu, Soon 
 Hwang, Junhyeok Park, and Seonghoon Ahn (Sogang University); JeoungAhn Par
 k, Jeongjin Lee, Jinna Yang, Soonyeal Yang, and Jungki Noh (SK hynix Inc.)
 ; Qing Zheng (Los Alamos National Laboratory (LANL)); Woosuk Chung and Hos
 hik Kim (SK hynix Inc.); and Youngjae Kim (Sogang University)\n-----------
 ----------\nDesign and Implementation of a Custom Hardware Accelerator for
  SZx Compression in Chipyard\n\nAbstract—Data movement bottlenecks have be
 come the dominant performance limiter in modern computing systems. At the 
 same time, scientific detectors generate overwhelming data\nvolumes; X-ray
  detectors may soon produce terabytes per second and high-energy physics e
 xperiments demand bandwidth on ...\n\n\nConnor Bohannon, Kazutomo Yohsii, 
 Sheng Di, Franck Cappello, and Antonino Miceli (Argonne National Laborator
 y (ANL))\n---------------------\nLunch break (on your own)\n\nRecording: L
 ivestreamed, Recorded\n\nRegistration Category: Technical Program Reg Pass
 , Workshop Reg Pass\n\nSession Chairs: Sheng Di (Argonne National Laborato
 ry (ANL), University of Chicago); Ana Gainaru (Oak Ridge National Laborato
 ry (ORNL)); Kento Sato (RIKEN Center for Computational Science (R-CCS)); X
 in Liang (University of Kentucky); and Jieyang Chen (University of Oregon)
END:VEVENT
END:VCALENDAR