First International Workshop on Performance Portable Programming Models for Accelerators (P^3MA)

Frankfurt, Germany
June 23, 2016


08:30 - 09:00a Welcome Coffee
09:00 - 09:05a Opening Remarks
(M. Graham Lopez, ORNL)
09:05 - 10:00a Keynote Address
(Si Hammond, SNL)
Session 1: Memory subsystem
10:00 - 10:30a "GPU-STREAM v2.0: Benchmarking the achievable memory bandwidth of many-core processors across diverse parallel programming models"
(Tom Deakin, University of Bristol)
10:30 - 11:00a "Battling Memory Requirements of Array Programming through Streaming"
(James Avery, University of Copenhagen, Denmark)
11:00 - 11:30a Coffee Break
Session 2: Parallel Programming using OpenMP and OpenACC
Session Chair: Simon McIntosh-Smith, University of Bristol
11:30 - 12:00p "Task-Based Cholesky Decomposition on Knights Corner using OpenMP"
(Jakub Kurzak, University of Tennessee)
12:00 - 12:30p "Software Cost Analysis of GPU-Accelerated Aeroacoustics Simulations in C++ with OpenACC"
(Marco Nicolini, RWTH Aachen University)
12:30 - 01:00p "From Describing to Prescribing Parallelism: Translating the SPEC ACCEL OpenACC Suite to OpenMP Target Directives"
(Guido Juckeland, HZDR, Germany)
01:00 - 02:00p Lunch
Session 3: Heterogeneous Computing
Session Chair: Guido Juckeland, HZDR, Germany
02:00 - 02:30p "A C++ Programming Model for Heterogeneous System Architecture"
(Ralph Potter, University of Bath and Codeplay Software Ltd.)
02:30 - 03:00p "Porting the MPI Parallelized LES Model PALM to Multi-GPU Systems - an Experience Report"
(Tobias Gronemeier, Institut für Meteorologie und Klimatologie, Leibniz Universität Hannover)
03:00 - 03:30p "Using C++ AMP to accelerate HPC applications on Multiple Platforms"
(M. Graham Lopez, ORNL)
03:30 - 03:55p Applications and OpenACC on Sunway TaihuLight
(Haohuan Fu, Deputy Director National Supercomputing Center in Wuxi)
03:55 - 04:00p Closing Remarks
(M. Graham Lopez)
04:00 - 04:30p Coffee Break