Schedule
09:00 PMBS Introduction and Welcome – America’s Center Convention Complex, St. Louis
Steven A. Wright
University of York, UK
Session 1: Performance Modeling
Chair: Steven A. Wright, University of York, UK
09:00 - 09:30
Modelling Load Imbalance In Shared Memory Multicore Systems [abstract]
Johannes Langguth
Simula Research Laboratory, Norway
University of Bergen, Norway
James Trotter
Simula Research Laboratory, Norway
Xing Cai
University of Oslo, Norway
Simula Research Laboratory, Norway
09:30 - 10:00
A Peak Performance Model for All-to-all on Hierarchical Systems and Its Applications [abstract]
Rohini Uma-Vaideswaran, Daniel Dotson, P. K. Yeung
Georgia Institute of Technology, USA
Joshua Romero, David Appelhans
NVIDIA Corporation, USA
10:00 - 10:30 Break
Session 2: Accuracy and Fidelity of Applications and Simulators
Chair: Sascha Hunold, TU Wien, Austria
10:30 - 11:00
Determining Levels of Detail for Simulators of Parallel and Distributed Computing Systems via Automated Calibration [abstract]
Jesse McDonald, Yick-Ching Wong, Henri Casanova
University of Hawaii at Manoa, USA
Kshitij Mehta, Frederic Suter, Rafael Ferreira Da Silva
Oak Ridge National Laboratory, USA
Loic Pottier
Lawrence Livermore National Laboratory, USA
Ewa Deelman
University of Southern California, USA
11:00 - 11:30 Best Paper
Beyond Guess and Check: Quantifying the Fidelity of Proxy Applications [abstract]
Si Chen
Emory University, USA
Simon Garcia de Gonzalo, Omar Aaziz, Jeanine Cook
Sandia National Laboratories, USA
Avani Wildani
Cloudflare, USA
Session 3: Short Papers
Chair: Murali Emani, Argonne National Laboratory, USA
11:30 - 11:45 Best Short Paper
CGSim: A Simulation Framework for Large Scale Distributed Computing Environment [abstract]
Sairam Sri Vatsavai, Kuan-Chieh Hsu, Ozgur Kilic, Yihui (Ray) Ren, David Park, Paul Nilsson, Sankha Dutta, Tasnuva Chowdhury, Adolfy Hoisie, Tadashi Maeno, Shinjae Yoo, Alexei Klimentov
Brookhaven National Laboratory, USA
Raees Khan Ahmed, Tania Korchuganova, Joseph Boudreau
University of Pittsburgh, USA
Shengyu Feng, Yiming Yang
Carnegie Mellon University, USA
Fatih Furkan Akman, Verena Ingrid Martinez Outschoorn, John Rembrandt (Remy) Steele
University of Massachusetts, USA
Scott Klasky, Norbert Podhorszki, Fred Suter
Oak Ridge National Laboratory, USA
Wei Yang
SLAC National Accelerator Laboratory, USA
11:45 - 12:00
PerfAnalyzer: Revealing Performance Trends using Version Oriented Visual Analysis of Scientific Software [abstract]
Kunal Pai, Mahyar Samani, Anusheel Nand, Jason Lowe-Power
University of California, Davis, USA
12:00 - 12:15
Implications of Full-System Modeling for Superconducting Architectures [abstract]
Sayef Azad Sakin
University of Utah, USA
Los Alamos National Laboratory, USA
James Ahrens
Los Alamos National Laboratory, USA
12:15 - 12:30
Experiences of Porting Structured and Unstructured Stencil Applications to FPGA using SYCL [abstract]
Zadok Storkey, Steven A. Wright, Ian Gray
University of York, UK
12:30 - 14:00 Lunch
Session 4: Large Language Model
Chair: Fred Suter, Oak Ridge National Laboratory, USA
14:00 - 14:30
MoE-Inference-Bench: Performance Evaluation of Mixture of Expert Large Language and Vision Models [abstract]
Krishna Teja Chitty-Venkata, Murali Emani, Venkatram Vishwanath
Argonne National Laboratory, USA
Sylvia Howland, Golara Azar, Daria Soboleva, Natalia Vassilieva
Cerebras, USA
Siddhisanket Raskar
Pacific Northwest National Laboratory, USA
14:30 - 15:00
Pretraining LLMs at Scale: Tuning Strategies and Performance Portability [abstract]
Adrián Pérez Diéguez, Àlex Batlle Casellas, Aleix Torres-Camps, Harris Teague, Jordi Ros-Giralt
Qualcomm, USA
15:00 - 15:30 Break
Session 5: Graphics Processing Units
Chair: TBA, TBA
15:30 - 16:00
Characterizing the Impact of GPU Power Management on an Exascale System [abstract]
Mariana Costa, Philippe O. A. Navaux, Arthur Lorenzon
Universidade Federal do Rio Grande do Sul, Brazil
Antigoni Georgiadou, James B. White III, Woong Shin, Bronson Messer
Oak Ridge National Laboratory, USA
Bruno Villasenor Alvarez, Jordà Polo
AMD, USA
16:00 - 16:30
A GPU FFT Wrapper to Co-optimize Floating-Point Precision and Library Selection via Predictive Error Modeling [abstract]
Julius Lehner, Eishi Arima, Martin Schulz
Technical University of Munich, Germany
Session 6: System Performance and Scheduling
Chair: Lilia Zaourar, CEA, France
16:30 - 17:00
ILAN: The Interference- and Locality-Aware NUMA Scheduler [abstract]
Edvin Mellberg, Axel Carlsson, Jing Chen, Miquel Pericas
Chalmers University of Technology, Sweden
17:00 - 17:30
On the Performance and Scalability of Cloud Supercomputers: Insights from Eagle and Reindeer [abstract]
Amirreza Rastegari, Prabhat Ram, Michael F. Ringenburg
Microsoft Corporation, USA
17:30 PMBS End