December 16, 2019, Sixth workshop in NIRICT GPGPU Reconnaissance workshop series
The workshop will feature invited talks about on GPU systems, analysis and applications.
Location: Vergadercentrum Vredenburg, Vredenburg 19, Utrecht
Registration: Everybody who is interested is welcome to attend, to listen to the talks, and to participate in discussions. There will not be a registration fee. However, for a smooth organisation, all participants are asked to register here.
Workshop Program:
10:00 - 10:30 Welcome with coffee
10:30 - 10:45 Workshop Introduction by Anton Wijs
10:45 - 11:15 Hanno Spreeuw (Netherlands eScience Center) - Calibration of observations from LOFAR and SKA on an NVIDIA Jetson Nano
11:15 - 11:45 Jelle van Dijk (University of Amsterdam) - Porting CUDA code to AMD GPU's using HIP
11:45 - 12:15 Mohsen Safari (University of Twente) - Formal Verification of Parallel Prefix Sum
12:15 - 13:30 Lunch
13:30 - 14:00 Muhammad Osama (Eindhoven University of Technology) - SIGmA: GPU Accelerated Simplification of SAT Formulas
14:00 - 14:30 Lars van den Haak (Eindhoven University of Technology) - Accelerating Nested Data Parallelism: Preserving Regularity
14:30 - 15:00 Bram Veenboer (ASTRON) - Radio-astronomical imaging: FPGAs vs. GPUs
15:00 - 15:30 Ben van Werkhoven (Netherlands eScience Center) - Preliminary results of auto-tuning GPU applications for energy efficiency using Kernel Tuner and PowerSensor
15:30 - 17:00 Drinks, Snacks & Networking


December 4, 2018, Fifth workshop in NIRICT GPGPU Reconnaissance workshop series
The workshop will feature invited talks about on GPU systems, analysis and applications.
Location: Vergadercentrum Vredenburg, Vredenburg 19, Utrecht
Registration: Everybody who is interested is welcome to attend, to listen to the talks, and to participate in discussions. There will not be a registration fee. However, for a smooth organisation, all participants are asked to register here.
Workshop Program:
10:00 - 10:30 Welcome with coffee
10:30 - 10:45 Workshop Introduction by Ben Van Werkhoven
10:45 - 11:15 Trevor McDonell (University of Utrecht) - A Functional Programming Language for GPUs
11:15 - 12:00 Valeriu Codreanu (SURFSara) - Design and Performance Evaluation of a Commodity GPU Cluster for HPC and Deep Learning Workloads
12:00 - 12:30 Henk Dreuning (University of Amsterdam) - A Beginner's Guide to Estimating and Improving Performance Portability
12:30 - 13:30 Lunch
13:30 - 14:00 Pieter Hijma (VU University Amsterdam) - Optimization Effectiveness: A Case-Study in Relating Performance to Programming Effort
14:00 - 14:30 Sagar Dolas (SURFsara) - Exploring the Potential of the ROCm Software Stack for High Performance Computing and Deep Learning on AMD GPUs
14:30 - 15:00 Maxwell Cai (Leiden University) - GPU-accelerated Research in Astrophysics
15:00 - 15:30 Ehsan Sharifi Esfahani (University of Amsterdam) - A survey on Energy Efficiency in GPUs
15:30 - 16:00 Merijn Verstraaten (Netherlands eScience Center) - Mix-and-Match: A Model-driven Runtime Optimisation Strategy for BFS on GPUs
16:00 - 17:00 Drinks, Snacks & Networking


May 16, 2018, Fourth workshop in NIRICT GPGPU Reconnaissance workshop series
The workshop will feature invited talks about on GPU systems, analysis and applications.
Location: Vergadercentrum Vredenburg, Vredenburg 19, Utrecht
Registration: Everybody who is interested is welcome to attend, to listen to the talks, and to participate in discussions. There will not be a registration fee. However, for a smooth organisation, all participants are asked to register here.
Workshop Program:
10:00 - 10:30 Welcome with coffee
10:30 - 11:00 Mohsen Safari - Parallel Algorithms to solve Shortest Path Problems in large graphs on GPUs
11:00 - 11:30 Pieter Hijma - An Ecosystem for Combining Performance and Correctness for Many-Cores
11:30 - 12:00 Roel Aaij - Potential for GPUs in HEP experiments: two examples
12:00 - 12:30 Shirin Dora - Deep Predictive Coding Networks on GPUs
12:30 - 13:30 Lunch
13:30 - 14:00 Antoine Veenstra - Edit distance on GPU clusters using MPI
14:00 - 14:30 Alfons Laarman - Optimal Storage of Combinatorial State Spaces
14:30 - 14:45 Coffee
14:45 - 15:15 Tanja Lange - Cryptanalysis using GPUs
15:15 - 15:45 Kai-Chun Ning - Solving large polynomial systems on GPUs
15:45 - 16:15 Muhammad Osama Mahmoud - SAT Simplifications on GPU Architecture
16:15 - 17:00 Brainstorming session: Roadmap for future Dutch GPU research
17:00 Drinks, Snacks & Networking


November 10, 2017, Third workshop in NIRICT GPGPU Reconnaissance workshop series
The workshop will feature invited talks about analysis of GPGPU applications.
Location: Vergadercentrum Vredenburg, Vredenburg 19, Utrecht
Registration: Everybody who is interested is welcome to attend, to listen to the talks, and to participate in discussions. There will not be a registration fee. However, for a smooth organisation, all participants are asked to register here.
Workshop Program:
10:00 - 10:30 Welcome with coffee
Analysing GPU code
10:30 - 11:00 Ben van Werkhoven (eScience Center) - Test-driven development of auto-tuned GPU code with Kernel Tuner
11:00 - 11:30 Marieke Huisman (University of Twente) - Verification of GPU kernels with VerCors
11:30 - 12:00 Alastair Donaldson (Imperial College London) - Exposing Errors Related to Weak Memory in GPU Applications
12:00 - 14:00 Lunch
Analysis with GPUs
14:00 - 14:30 Dragan Bosnacki (Eindhoven University of Technology) - Probabilistic Model Checking on GPUs
14:30 - 15:00 Qixia Yuan (U. Luxembourg) - GPU-accelerated Steady-state Computation of Large Probabilistic Boolean Networks
15:00 - 15:30 Coffee
15:30 - 16:00 Thomas Neele (Eindhoven University of Technology) - Optimizations in GPU model checking
16:00 - 16:30 Anton Wijs (Eindhoven University of Technology) - Efficient GPU algorithms for parallel decomposition of graphs into strongly connected and maximal end components
16:30 - 17:10 Brainstorming session: A Dutch GPU community
17:10 Drinks, Snacks & Networking


June 8, 2017, Second workshop in NIRICT GPGPU Reconnaissance workshop series
The workshop will feature invited talks about interesting GPGPU applications in different domains, from both industry and academia.
Location: Vergadercentrum, Vredenburg 19, Utrecht
Registration: All interested individuals are welcome to attend, to listen to the talks, and to participate in discussions. There will not be a registration fee charged to participants. However, all attendees must register.
Workshop Programme:
9:45 welcome with coffee
10:00 Ben van Werkhoven (Netherlands eScience Center) - GPU Computing at the Netherlands eScience Center
10:30 Dragan Bosnacki (TU/e) - Inference of gene regulatory networks on GPUs
11:00 coffee
11:15 Jeroen Bedorf (Sterrenwacht, U. Leiden) - GPU Usage in Gravitational Dynamics
11:45 Vincent Hindriksen (StreamComputing)- OpenCL: Industry Examples and 3.0 Preview
12:15 lunch
14:00 Cedric Nugteren (TomTom) - CLBlast: A Tuned BLAS Library for Faster Deep Learning
14:30 Willem Jan Palenstijn (CWI)- Using GPUs in Tomography
15:00 Andrei Jalba (TU/e)- GPU computing for scientific visualization and computer graphics
15:30 coffee
16:00 Sven Warris (U. Wageningen) - Speeding up sequence alignment enables new DNA sequence approaches
16:30 Bram Veenboer (ASTRON) - Efficient Imaging in Radio Astronomy Using GPUs
17:00 closing and drinks



October 28, 2016, First workshop in NIRICT GPGPU Reconnaissance workshop series on
"GPGPU Systems: from hardware to programming and performance"
Location:Vergadercentrum, Vredenburg 19, Utrecht
Call for Contributions.

Workshop Programme
9:30-10:00 Coffee and greeting(s)
10:00-10:05 Welcome
10:05-10:45 Kernel Tuner: A simple CUDA/OpenCL kernel tuner in Python - Ben van Werkhoven, Berend Weel and Hanno Spreeuw
10:45-11:25 Analyzing the Complexity of Auto-Tuning Many-Core Accelerators - Alessio Sclocco, Henri E. Bal and Rob V. van Nieuwpoort
11:25-12:05 Programming Many-Cores on Multiple Levels of Abstraction - Pieter Hijma
12:10-13:00 Lunch
13:00-13:40 Multi-GPU code generation using Bones - Gert-Jan van den Braak, Roel Jordans and Henk Corporaal
13:40-14:20 A Scalable GPU Accelerated Framework for Risk-Neutral Model Calibration and (Nested) Simulation - Alex de Geus
14:20-15:00 High-Performance Graph Processing on Heterogeneous CPU-GPU Platforms using Dynamic Load-Balancing - Stijn Heldens, Ana Lucia Varbanescu and Alexandru Iosup
15:00-15:30 Performance Prediction and Bottleneck Analysis for GPGPU Applications - Souley Madougou, Ana Lucia Varbanescu
15:30-15:45 Coffee break
15:45-16:30 Brainstorming session: GPGPU systems
16:30-17:00 Drinks, Snacks, & Networking



March 22, 2016, NIRICT GPU in ICTOpen 2016
Marieke Huisman (University of Twente): NIRICT GPU reconnaissance project.



December 3, 2015, Kickoff Meeting
Location: Vergadercentrum Vredenburg (Vredenburg 19), Utrecht
Workshop Programme
10:00 opening & welcome - goal of the day and the reconnaissance theme - Marieke Huisman
10:05 Verification of OpenCL programs - Marieke Huisman
10:20 Alexandru Iosup
10:40 Andrei Jalba
11:00 GPU-related work in the CAES group - Jan Kuper
11:20 Fast Solvers for Linear Systems on the GPU - Kees Vuik
11:40 GPU research in the Electronic System group at the University of Eindhoven - Gert-Jan van den Braak
12:00 lunch
14:00 Ana Lucia Varbanescu
14:20 Anton Wijs - GPU Accelerated Verification and Verified Model-Based Development of Multi-Threaded Software
14:40 coffee break
15:00 - 17:00 discussion