Nvidia TESLA-C1060 card (240 CUDA cores)

Co-ordinator: Prof. Harry Varvoglis (AUTh)

Main Research Group 1

Harry Varvoglis (AUTh - Coordinator)

George Voyatzis (AUTh, cv)

Kleomenis Tsiganis (AUTh, cv)

Michael Vrahatis (UPAT, cv)

Main Research Group 3

Panos Patsis (RCAAM, Co-I, cv)

Christos Efthymiopoulos (RCAAM, cv)

Anastasios Anastasiadis (ISARS-NOA, cv)

Costis Gontikakis (RCAAM, cv)

GPU4PHYS:

Studying complex problems in Computational Physics through the development of GPU-based software

Proposal submitted in call THALIS (2010, Ministry of Education, Greece)

web_page: www.astro.auth.gr/~varvogli

( click here for my CV )

Main Research Group 2

Nikos Stergioulas (AUTh, Co-I, cv)

Kostas Kokkotas (AUTh, cv)

Manolis Plionis (IAA-NOA, cv)

Spyros Vasilakos (RCAAM, cv)

Invited Researcher

David Kaufmann (SwRI, cv)

External Colaborators

Theodore Simos (Un. Peloponese, cv)

Harold Levison (SwRI, USA, cv)

Harris Skokos (MPIKS, cv)

Brief Description of the Project

Computational Physics consists in using numerical methods for studying complex physical problems, typically through massive simulations. Many “hot” problems, such as the formation and evolution of galaxies or planetary systems, require immense computing power, not always available (or extremely costly to develop). This may be overcome with the use of general-purpose Graphics Processing Units (gp-GPUs, a.k.a. programmable graphics cards). New-generation GPUs can turn a normal PC to a small super-computer at a very low cost.

We intend to exploit GPU technology to study a wide range of problems in computational physics, from stability of charged particle beams to evolution of cosmological structures. We will upgrade an existing GPU-based platform (at AUTh: 2 Quad-core PCs equipped with 4x and 2x (resp.) Nvidia Tesla-C1060 GPU boards – each with a ~1 Tflops performance for single-precision arithmetics), and develop the necessary software. Our team is composed of experienced researchers from different fields, all with long experience in scientific software development and computational methods.

Expected Results

Development of innovative research infrastructure (GPU-based platform, library of scientific software)

Training of young researchers. Development and diffusion of educational material to Graduate Students and general public.

Establishment of links with the local software industry.

We expect to achieve ground-braking scientific results on the following research topics:

trajectory desing in astrodynamics, using advanced phase-space exploration tools

combined with intelligent optimization techniques

formation and evolution of the Solar System and Extrasolar Planetary Systems, using

GPU-based N - body and N - body/SPH codes

formation and evolution of Large-Scale structures in the Universe (galaxies, clusters, etc.),

using GPU-based smoothed-potential, N - body and N - body/hydro codes

gravitational radiation from Collapsing Binaries in clusters and Magnetized Neutron Stars,

using GPU-based N – body (post-newtonian) and relativistic MHD codes

the role of Particle-Particle Interactions in Space Plasma and Accelerator Beam Dynamics,

using GPU-based, Particle-in-Cell and Coulomb-N-body codes