Extended Version of Parallel Processing Conf. 1992 Paper and Revised Version of Stony Brook Tr # 92/01 Eager Sharing for Eecient Massive Parallelism

{ Workstation networks can become tera-FLOPS supercomputers by adding highspeed interfaces supporting selective eager sharing. For Gaussian elimination and fast Fourier transform, selective eager sharing is much more eecient than global sharing of all data changes, and average eeciency remains above 60% for thousands of processors. Prototype SESAME interfaces will share data at 50 megabytes/second among more than 100 workstations. Propagation delays are typically 0:8 microseconds and overlap computations. All shared data reads are quick local accesses. Eager sharing supports diiuse non-local accesses in ne-grained parallel programs much more eeciently than demand driven cache protocols. Future massively parallel supercomputers should ooer eagersharing coherence mechanisms.