Session 1 – Best Paper Candidates
Wednesday, Nov. 2nd — 14h30 – 16h00
Chair: Viktor Prasana (USC, USA)
- Pangbo Sun, Hao Wu, Jiangming Jin, Ziyue Jiang, and Yifan Gong
TCUDA: A QoS-based GPU Sharing Framework for Autonomous Navigation Systems - Hammurabi Mendes, Bryce Wiedenbeck and Aidan O’Neill
Seriema: RDMA-based Remote Invocation with a Case-Study on Monte-Carlo Tree Search - Elvis Rojas, Diego Pérez and Esteban Meneses
Exploring the Effects of Silent Data Corruption in Distributed Deep Learning Training - Erhan Tezcan, Tugba Torun, Fahrican Koşar, Kamer Kaya, and Didem Unat
Mixed and Multi-Precision SpMV for GPUs with Row-wise Precision Selection
Session 2 – Memory Systems
Wednesday, Nov. 2nd — 16h30 – 18h30
Chair: José Nelson Amaral (University of Alberta, Canada)
- João Vieira, Nuno Roma, Gabriel Falcao and Pedro Tomás
gem5-ndp: Near-Data Processing Architecture Simulation From Low Level Caches to DRAM - João Fabrício Filho, Isaías Felzmann, and Lucas Wanner
Approximate Memory with Protected Static Allocation - Brady Testa, Samira Mirbagher-Ajorpaz, and Daniel A. Jiménez
Dynamic Set Stealing to Improve Cache Performance - Arthur M. Krause, Paulo C. Santos, and Philippe O. A. Navaux
Avoiding Unnecessary Caching with History-Based Preemptive Bypassing - Alex Weaver, Krishna Kavi, Pranathi Vasireddy and Gayatri Mehta
Memory-Side Acceleration and Sparse Compression for Quantized Packed Convolutions - Sandra Catalán, Francisco D. Igual, Rafael Rodríguez-Sanchez, José R. Herrero, and Enrique S. Quintana-Ortí
NUMA-Aware Dense Matrix Factorizations and Inversion with Look-Ahead on Multicore Processors
Session 3 – Parallel Algorithms and Applications
Thursday, November 3rd — 08h30 – 10h30
Chair: Hermes Senger (UfSCar, Brazil)
- Alexander van der Grinten, Geert Custers, Duy Le Thanh and Henning Meyerhenke
An MPI-Parallel Algorithm for Static and Dynamic Top-k Harmonic Centrality - Samuel Ferraz, Vinicius Dias, Carlos H. C. Teixeira, George Teodoro and Wagner Meira Jr.
Efficient Strategies for Graph Pattern Mining Algorithms on GPUs - Daniel Wladdimiro, Luciana Arantes, Pierre Sens and Nicolas Hidalgo
A predictive approach for dynamic replication of operators in distributed stream processing systems - James Almgren-Bell, Nader Al Awar, Dilip S. Geethakrishnan, Milos Grigoric, and George Biros
A Multi-GPU Python Solver for Low-Temperature Non-Equilibrium Plasmas - Samuel Cajahuaringa, Leandro N. Zanotto, Daniel L. Z. Caetano, Sandro Rigo, Hervé Yviquel, Munir S. Skaf and Guido Araujo
Ion-Molecule Collision Cross-Section Simulation using Linked-cell and Trajectory Parallelization - Javier Garcia-Blas, Javier Fernandez Muñoz, Jesus Carretero, Fabrizio Marozzo, Domenico Talia, Paolo Trunfio, Alberto Fernandez-Pena, and Daniel Martin de Blas
Convergence of HPC and Big Data in extreme-scale data analysis through the DCEx programming model
Session 4 – Computer Architecture
Thursday, November 3rd — 13h30 – 14h30
Chair: Edson Borin (UNICAMP, Brazil)
- Manuel F. Dolz, Héctor Martínez, Pedro Alonso-Jorda and Enrique S. Quintana-Orti
Convolution Operators for Deep Learning Inference on the Fujitsu A64FX Processor - Guillaume Didier, Clémentine Maurice, Antoine Geimer and Walid J. Ghandour
Characterizing Prefetchers using CacheObserver - Fareed Qararyah, Muhammad Waqar Azhar and Pedro Trancoso
FiBHA: Fixed Budget Hybrid CNN Accelerator
Session 5 – Energy Consumption
Friday, November 4th — 08h30 – 09h30
Chair: Gerald F. Lofstead (Sandia National Laboratories)
- Thierry Arrabal, Lucas Betencourt, Eddy Caron and Laurent Lefevre
Setting up an experimental framework for immersion cooling system and analysis - Jonathas Silveira, Lucas Castro, Victor Araújo, Rodrigo Zeli, Daniel Lazari, Marcelo Guedes, Rodolfo Azevedo and Lucas Wanner
Prof5: A RISC-V profiler tool - Emmanuel Agullo, Marek Felšöci, Amina Guermouche, Hervé Mathieu, Guillaume Sylvand and Bastien Tagliaro
Study of the processor and memory power consumption of coupled sparse/dense solvers
Session 6 – Performance Evaluation
Friday, November 4th — 09h30 – 10h30
Chair: Gerald F. Lofstead (Sandia National Laboratories)
- Aravind Sankaran and Paolo Bientinesi
A Test for FLOPs as a Discriminant for Linear Algebra Algorithms - Miguel G. Xavier, Carlos H. C.Cano, Vinícius Meyer, and César A. F. De Rose
IntP: Quantifying cross-application interference via system-level instrumentation - Alexander V. Goponenko, Kenneth Lamar, Christina Peterson, Benjamin A. Allan, Jim M. Brandt and Damian Dechev
Metrics for Packing Efficiency and Fairness of HPC Cluster Batch Job Scheduling
Session 7 – Cloud Computing
Friday, November 4th — 13h30 – 14h30
Chair: Cristina Boeres (UFF, Brazil)
- Rafaela C. Brum, Pierre Sens, Luciana Arantes, Maria Clicia Castro, and Lúcia Maria de A. Drummond
Optimizing Execution Time and Costs of Cross-Silo Federated Learning Applications with Datasets on different Cloud Providers - Vanderlei Munhoz, Márcio Castro and Odorico Mendizabal
Strategies for Fault-Tolerant Tightly-coupled HPC Workloads Running on Low-Budget Spot Cloud Infrastructures - Maxim Moraru, Adrien Roussel, Hugo Taboada, Christophe Jaillet, Marc Perache and Michael Krajecki
Performance improvements of parallel applications thanks to MPI-4.0 hints
Session 8 – Parallel I/O and Big Data
Friday, November 4th — 14h30 – 15h30
Chair: Shadi Ibrahim (Inria Rennes, France)
- Yang Chen, Feng Zhang, Yinhao Hong, Yunpeng Chai, Wei Lu, Hong Chen, Xiaoyong Du, Peipei Wang, Le Mi, Jintao Li, Xilin Tang, Yanliang Zhou, Wei Zhou, Peng Zhang, Fengyi Chen, Pengfei Li, Yu Li
Taming the Big Data Monster: Managing Petabytes of Data with Multi-Model Databases - Matheus Tavares Bernardino and Alfredo Goldman
Parallelizing Git Checkout: a Case Study of I/O Parallelism - Igor Fontana de Nardin, Patricia Stolf and Stephane Caux
Analyzing Power Decisions in Data Center Powered by Renewable Sources
Session 9 – Scheduling
Friday, November 4th — 16h00 – 17h00
Chair: Mathieu Faverge (Inria Bordeaux, France)
- Omar Shaaban, Jimmy Aguilar, Vicenç Beltran, Paul Carpenter, Eduard Ayguadé and Jesus Labarta Mancho
Automatic aggregation of subtask accesses for nested OpenMP-style tasks - Jing Chen, Madhavan Manivannan, Bhavishya Goel, Mustafa Abduljabbar and Miquel Pericàs
STEER: Asymmetry-aware Energy Efficient Task Scheduler for Cluster-based Multicore Architectures - Odin Ugedal and Rakesh Kumar
Mitigating Unnecessary Throttling in Linux CFS Bandwidth Control
