Future Generation Computer Systems 15 (1999) 807–816 Metacomputing experience in a transatlantic wide area application test-bed Michael M. Resch * , Dirk Rantzau, Robert Stoy High Performance Computing Center Stuttgart (HLRS), Allmandring 30, D-70550 Stuttgart, Germany Accepted 14 December 1998 Abstract In the frame of a G7 initiative the High Performance Computing Center Stuttgart (HLRS) together with the Pittsburgh Supercomputing Center (PSC) and Sandia National Laboratories (SNL) has set up a transatlantic wide area application test-bed in 1997. A dedicated ATM-Link was installed that connected German research networks to vBNS and ESnet. During 1 year this test-bed was extensively used for metacomputing and collaborative working. Two applications – one from computational fluid dynamics and one from molecular dynamics – were adapted and run on the test-bed. For message-passing an MPI library was implemented that supports metacomputing. An already existing software for collaborative visualization was adapted for that scenario. This article describes the technical background of the cooperation, results that have been achieved for the two applications so far and lessons that have been learned. Special emphasis will be given to future work planned. ©1999 Elsevier Science B.V. All rights reserved. 1. Introduction In the frame of the G7 Global Information Society Initiative “Global Interoperability of Broadband Net- works”, two proposals for a transatlantic co-operation were submitted. One was from Pittsburgh Supercom- puting Center (PSC) and High Performance Comput- ing Center Stuttgart (HLRS) and was focusing on the application aspect of metacomputing. The other one was from Sandia National Laboratories (SNL) and HLRS and was concentrating on distributed visual- ization in a virtual laboratory. During the first project phase it became clear that the projects should be merged into a Global Wide Area Application Test-bed (G-WAAT). This would allow to couple simulation and visualization in a metacomputing scenario. * Corresponding author The main targets of the merger were: • To set up a production test-bed for metacomputing applications and distributed visualization. • To combine supercomputing forces in order to solve much larger problems than any of the partners could solve on his own resources. • To integrate software components in order to es- tablish a collaborative simulation and visualization environment. In a first step this meant to set up a network connec- tion fast enough to allow distributed simulation and visualization. Second, it was necessary to find a com- munication software that enables metacomputing for one single application. Third, distributed visualization software had to be adapted and extended. In response to these needs a transatlantic network connection was set up. The communication issue was resolved by implementing a completely new communication library based on the MPI standard. 0167-739X/99/$ – see front matter ©1999 Elsevier Science B.V. All rights reserved. PII:S0167-739X(99)00028-X