PUMPS Summer School, June 26-30, 2017

NVIDIA kindly donated 2 Titan X GPUs for the best poster and best clinic awards and 1 Jetson TX1 for the best achievement award.

Best Poster Awards

Title: Exploiting Data-Parallelism in ILUPACK using Graphics Processors
Author(s): José I. Aliaga, Mathias Bollhöfer, Ernesto Dufrechou, Pablo Ezzatti, Enrique Quintana-Ortí
Affiliation: Universitat Jaume I (Spain), TU-Braunschweig (Germany), Universitat de la República (Uruguay)
Poster: Exploiting Data-Parallelism in ILUPACK using Graphics Processors

The solution of sparse linear systems of large dimension is a critical step in problems that arise in a diverse range of applications. For this reason, a number of iterative solvers have been developed, among which ILUPACK stands out due to an inverse-based multilevel ILU preconditioner with appealing numerical properties. We have tried to enhance the computational performance of ILUPACK by off-loading the execution of several key computational kernels to one or several Graphics Processing Unit (GPU). In particular, we have targeted the preconditioned CG method for symmetric linear systems, Gmres and Bicg methods for sparse general systems and the preconditioned Sqmr method for sparse symmetric indefinite problems in ILUPACK. We have also included GPU computing in the task-prallel version of ILUPACK, both for distributed and shared memory systems.

Best Clinic Award

Title: CUDA Accelerated Wavelet Neural Network for Time Series Forecasting
Author: Anton Lebedev
Affiliation: Universität Tübingen (Germany)
The code I would like to parallelize on the GPU implements an Runge-Kutta type integrator scheme for a stochastic Schrödinger equation. Whilst the scheme itself is sequential the stoch. Schrödinger eqn. has to be solved for multiple (1000+) sample paths which are independent of each other.

Best Achievement Award

PUMPS 2017 best student: Martin Biel

awards.txt · Last modified: 2017/07/04 18:20 by pfarre
Drupal Garland Theme for Dokuwiki