Performance Modeling, Benchmarking and Simulation 
of High Performance Computer Systems (PMBS19) 

held as part of ACM/IEEE Supercomputing 2019 (SC19), Denver, CO, USA

10th IEEE International Workshop on

Welcome to PMBS 2019
18th November 2019
Location: 603

9:00 - 9:10
A Decade of the Performance Modeling, Benchmarking and Simulation Workshop

Stephen Jarvis
University of Warwick, UK

Session 1: Best Papers
Chair: Stephen Jarvis

9:10 - 9:30
Automatic Throughput and Critical Path Analysis of x86 and ARM Assembly Kernels
  [abstract] [paper] [presentation]

Jan Laukemann, Julian Hammer, Georg Hager, Gerhard Wellein
University of Erlangen-Nuremberg, Germany

9:30 - 10:00
An Instruction Roofline Model for GPUs
  [abstract] [paper] [presentation]

Nan Ding, Samuel Williams
Lawrence Berkeley National Laboratory, CA

10:00 - 10:30 Coffee Break

Session 2: GPUs for Monte Carlo
Chair: Sebastiano Fabio Shifano

10:30 - 11:00
Exploiting Hardware-Accelerated Ray Tracing for Monte Carlo Particle Transport with OpenMC
  [abstract] [paper] [presentation]

Justin Salmon, Simon McIntosh-Smith
University of Bristol, UK

11:00 - 11:30
Enhancing Monte Carlo proxy applications on GPUs
  [abstract] [paper]

Forrest Shriver, Justin Watson
University of Florida, FL

Seyong Lee, Steven Hamilton, Jeffrey Vetter
Oak Ridge National Laboratory, TN

Session 3: Late-breaking Research and Preliminary Techniques
Chair: Misbah Mubarak

11:30 - 11:50
Comparing Managed Memory and ATS with and without Prefetching on NVIDIA Volta GPUs
  [abstract] [paper] [presentation]

Rahulkumar Gayatri, Kevin Gott, Jack Deslippe
NERSC, Lawrence Berkeley National Laboratory, CA

11:50 - 12:10
Testing the Limits of Tapered Fat Tree Networks
  [abstract] [paper]

Philip A. Taffet
Rice University, TX

Sanil Rao
Carnegie Mellon University, PA

Edgar A. León, Ian Karlin
Lawrence Livermore National Laboratory, CA

12:10 - 12:30
Validation of gem5 for x86 Platforms
  [abstract] [paper] [presentation]

Ayaz Akram
University of California, Davis, CA

Lina Sawalha
Western Michigan University, MI

12:30 - 14:00 Lunch

Session 4: Performance Profiling and Monitoring
Chair: Stephen Jarvis

14:00 - 14:30
A generalized statistics-based model for predicting network-induced variability
  [abstract] [paper] [presentation]

Sudheer Chunduri, Elise Jennings, Kevin Harms, Christopher Knight, Scott Parker
Argonne National Laboratory, IL

14:30 - 15:00
CUDA Flux: A Lightweight Instruction Profiler for CUDA Applications
  [abstract] [paper] [presentation]

Lorenz Braun, Holger Fröning
Heidelberg University, Germany

15:00 - 15:30 Coffee Break

Session 5: Performance Analysis
Chair: Ian Karlin

15:30 - 16:00
OMB-UM: Design, Implementation, and Evaluation of CUDA Unified Memory Aware MPI Benchmarks
  [abstract] [paper]

Karthik Vadambacheri Manian, Ching-Hsiang Chu, Ammar Ahmad Awan, Kawthar Shafie Khorassani, Hari Subramoni, Dhabaleswar K. Panda
Ohio State University, OH

16:00 - 16:30
Fine-Grained Analysis of Communication Similarity between Real and Proxy Applications
  [abstract] [paper]

Omar Aaziz, Courtenay Vaughan, Jeanine Cook
Sandia National Laboratories, NM

Jonathan Cook
New Mexico State University, NM

Jeffery Kuehn
Los Alamos National Laboratory, NM

David Richards
Lawrence Livermore National Laboratory, CA

16:30 - 17:00
Performance Analysis of Deep Learning Workloads on Leading-edge Systems
  [abstract] [paper] [presentation]

Yihui Ren, Shinjae Yoo, Adolfy Hoisie
Brookhaven National Laboratory, NY

Program Committee
Submit Paper