At the University of Maryland, our group has a network of Sun and Digital Alpha workstations, connected to the Internet.
Within the Department of Computer Science and UMIACS, we have:
The parallel machines, and many workstations, are networked via ATM OC-3 connections to other departments on the Maryland campus, and Maryland has also recently acquired an NSF vBNS ATM connection to the NSF supercomputing centers and other sites.
Rutgers has a high performance 8-node UltraSPARC cluster with Myrinet Switch, an 8-node IBM SP-2, a 512-node nCUBE2s, and a large heterogeneous workstation farm connected with Ethernet(Alpha, SGI, SUN and HP). Rutgers also has access to a large CRAY-T3E and IBM SP-2.
The University of California at Santa Barbara (subcontractors to Rutgers) has a 64-processor Meiko CS-2 distributed-memory machine supported by a major NSF CISE infrastructure grant, an SCI-based SUN UltraSPARC-1/2 workstation cluster supported by an NSF instrumentation grant along with matching funds from UC and SUN MICRO, a 4-node SGI CHALLENGE DM and a 32-node SGI ORIGIN2000 shared memory system supported by another NSF infrastructure grant. We have access to the Cray T3D/E at the San Diego Super Computing Center and an HP/Convex 32-node Exemplar shared memory machine at Navy NRAD.
Syracuse has a 12 node IBM SP-2, 8-node Sun UltraSPARC cluster with ATM connectivity and a 12 node Windows NT PC cluster with fast Ethernet connection. Syracuse also has an 8 node Silicon Graphics Power Challenge. We have excellent ATM networking with NSF vBNS as well a NYNEX experimental ATM link (NYNET). Syracuse systems are optimized for I/O and database support as part of our New York State funded InfoMall Industry outreach activity. Syracuse uses facilities at NSF centers for large scale computations such as those in the numerical black hole project.
Through the Internet, we have access to large parallel machines at several sites. Among these machines are: