Convegno
Perspectives of GPU Computing
in Physics and Astrophysics
lunedì 15 settembre 2014 - mercoledì, 17 settembre 2014

Dipartimento di Fisica
piazzale Aldo Moro 5
00185, Roma

Conference organized by Department of Physics Sapienza, IAC, INFN, INGV, Roma, Italy
Sponsored also by: Euretile, Harpa-HP, Physics at the Terascale, nVIDIA

SCIENTIFIC PROGRAM

MONDAY, SEPT. 15, 2014

09:15– 09:30 Welcoming Address

Science with GPUs:

First Morning Session

09:30–10:00 Deborah Joanne BARD (SLAC Natl. Acc. Lab., Stanford, USA)
Cosmological Calculations on the GPU

10:00–10:20 Claudio GHELLER (ETH, CSCS, Switzerland)
Ramses for GPU Accelerated Numerical Cosmology

10:20–10:40 Mario SPERA (INAF, Astr. Obs. of Padova, Italy)
Using GPUs to Solve the Classical N-Body Problem In Physics and Astrophysics

10:40–11:00 Yohai MEIRON (KIAA-PKU, China)
Expansion Techniques for Stellar Dynamics on GPUs

11:00–11:30 Coffee Break

Second Morning Session

11:30–11:50 Long WANG (Dep. Of Astronomy, Univ. of Beijing, China)
Acceleration of The Direct N-Body NBODY6++ With GPU

11:50–12:10 Michela MAPELLI (INAF, Astr. Obs. of Padova, Italy)
Simulating the Nursery of Stars with GPUs

12:10–12:30 Simon GRIMM (Univ. Zurich, Switzerland)
GENGA: a GPU code for Planet Formation and Evolution

12:30–15:00 Lunch Time

First Afternoon Session

15:00–15:30 Jeroen BÉDORF (Univ. of Leiden, The Netherlands)
Galaxy Simulations Powered by GPUs

15:30–15:50 Denis PERRET (LESIA, Obs. de Meudon, France)
Design of a GPU Based RTC for E-ELTs Adaptive Optics

15:50–16:10 Gary DEBRECZENI (VIRGO and Res. Inst. for P and N. Physics, Budapest, Hungary)
Gravitational Wave Physics is in The Doorstep of a New, Very Exciting Era

16:10–16:30 Karel ADÁMEK (Silesian Univ. in Opava, Czech Republic)
The Implementation and Comparison of a Polyphase Filter on Many–CoreSystems

16:30–17:00
Coffee Break

Second Afternoon Session

17:00–17:30 Thomas SCHULTHESS (ETH, Zurich, Switzerland)
Performance Portable Application Software Strategies and Energy Efficiency Considerations

17:30–17:50 Manuel J. CASTRO DíAZ (Univ. of Malaga, Spain)
HySEA-Tsunami Model: a GPU Implementation for the Italian TEWS

17:50–18:20 Dimitri KOMATITSCH (CNRS, Marseille, France)
Exascale Challenges for Acoustic Waves and Inversion

18:20–18:50 Peter MESSMER (ETH-nVIDIA, Zurich, Switzerland)
GPU Optimization for Seismic Applications

TUESDAY, SEPT. 16, 2014

The international context:

First Morning Session

09:30–10:00 Sergi GIRONA (PRACE, Barcelona, Spain)
PRACE: access to Tier-0 systems, and enabling the access to ExaScale systems

10:00–10:20 Rainer SPURZEM (Natl. Astron. Observatories, Chinese Academy, China)
Astrophysical GPU Supercomputing in China and Europe

10:20–10:50 Massimiliano FATICA (nVIDIA, S.ta Clara, USA)
GPU Computing for Scientific Applications: Past, Present and Future

10:50–11:20 Carlo CAVAZZONI (CINECA, Italy)
Exascale: challenges and opportunities in a power constrained world

11:20–11:50

Coffee Break

Science with GPUs:

Second Morning Session

11:50–12:20 Dirk PLEITER (Univ. of Regensburg and Exasc. Inn. Cent., Julich, Germany)
GPUs in Physics: Analysis of Architectural Features

12:20–12:40 Dániel BERÉNYI (Wigner RCP, Budapest, Hungary)
Code Generation for Differential equation Solvers

12:40–13:00 Ferenc NAGY EGRI (HAS Wigner RCP, Budapest, Hungary)
Static GPU Code Generation in Scientific Computing

13:00–15:00 Lunch Time
First Afternoon Session

15:00–15:30 Joshua ANDERSON (Univ. of Michigan, USA)
GPU Accelerated Soft Matter Particle Simulations

15:30–16:00 Axel KOHLMEYER(Temple Univ., USA)
After the Gold-Rush: How Have GPUs Changed Condensed Matter Simulations?

16:00–16:20 Lorenzo ROVIGATTI (Univ. of Vienna, Austria)
Investigating the Phase Behaviour of Valence-Limited DNA Constructs on GPUs

16:20–16:40 Vladimir KUKULIN (Moscow State Univ., Russia)
New Solutions for Quantum Scattering Problems via GPU

16:40–17:10 Coffee Break

Second Afternoon Session

17:10–17:40 Simone MELCHIONNA (IPCF, CNR, Roma, Italy)
Simulations of Proteins Suspensions in Crowding Conditions on Massive GPU Clusters

17:40–18:00 Marco BAITY JESI (Univ. Complutense, Madrid, Spain)
Phase Transition in Heisenberg Spin Glasses

18:00–18:20 Miguel BERGANZA (Sapienza Univ. of Roma, Italy)
Parallel Monte Carlo Simulations of Spin Models in Non-Regular Topologies

18:20–18:40 Matteo LULLI (Sapienza Univ. of Roma, Italy)
3D Ising Spin Models:a Test Ground for Wide Spectrum Optimizations

WEDNESDAY, SEPT. 17, 2014

Science with GPUs:

First Morning Session

09:30–10:00 Alessandro LONARDO (INFN, Roma, Italy)
A FPGA–Based Network Interface Card with GPUDirect Enabling Real–Time GPU Computing in HEP Experiments

10:00–10:20 Lorenzo RINALDI (Univ. of Bologna and INFN, Bologna, Italy)
GPGPU for Track Finding and Triggering

10:20–10:40 Francesco ROSSI (Univ. of Bologna and INFN, Bologna, Italy)
Robust Algorithms for Current Deposition and Dynamic Load-balancing in a GPU Particle–in–Cell Code

10:40–11:00 Peter HOBSON (Brunel University, Uxbridge, UK)
Many–Body Simulations Based on Ball Trees and Separated Pair Decompositions

11:00–11:30 Coffee Break

Some Hardware Perspectives:

Second Morning Session

11:30–12:00 Alberto GALLI (Hewlett Packard, Italy)
Accelerators Technology Update

12:00–12:30 Christopher DAHNKEN (INTEL, SSG DRD EMEA, USA)
Intel Xeon Phi: How to get the most out of the Intel Co-Processor

12:30–13:00 Piero ALTOÈ (E4 Computer Engineering, Italy)
HPC Benchmarks on ARM64+GPUs

13:00–15:00 Lunch Time

Science with GPUs:

First Afternoon Session

15:00–15:30 Davide ROSSETTI (INFN, Roma, Italy)
Efficient Data Movement on GPU-Accelerated Clusters

15:30–15:50 Bronson MESSER (Oak Ridge Natl. Laboratory, USA)
Moving Applications to Titan

15:50–16:10 Mauro GAROFALO (Univ. of Napoli Federico II, Italy)
Acceleration of Machine Learning Models Based on GPU

16:10–16:30 Francesco SIMULA (INFN, Roma, Italy)
Distributed Simulations of Polychronous and Plastic Spiking Neural Networks: Experiment with GPUs

16:30–17:00 Coffee Break

Second Afternoon Session

17:00–17:20 Angelo SCHIAVI (SBAI, Sapienza Univ. of Roma, Italy)
Monte Carlo treatment Planning in Particle Therapy Using Multi-Core Multi–GPU Hardware

17:20–17:40 Enrico CALORE (INFN, Ferrara, Italy)
Design and Optimization of a Portable Lattice Boltzmann Code for Heterogeneous Architectures

17:40–18:00 Sam THOMSON (Univ. of Edinburgh, UK)
Taranis: Ray Tracing Radiative Transfer in SPH

18:00–18:30 Meeting Final Discussion and Conclusion


Per maggiori informazioni clicca qui


Informativa      Aracneeditrice.it si avvale di cookie, anche di terze parti, per offrirti il migliore servizio possibile. Cliccando 'Accetto' o continuando la navigazione ne acconsenti l'utilizzo. Per saperne di più
Accetto