We propole a machine architecture for a high-performaxe proceuieg node for a mueage-pas&g, MIMD concurrent computer. The principal mee,haaismn for attaining this goal are the direct execution and bufferius of meesylee and a memory-based architecture that permit@ very fast contut witches. Our architecture &o ineluder a aevel msmory orga-nis&x~ that permita both indexed and aeeecikve acceuu and that incorporates sa itwtruction buffer and me-age queer. Simulation re mite suggest that thL architecture reduccl mauge reception overheed by more thau an order of magnitude.
No takes yet. Share an insight, caveat, or question.
Dally et al. (1998) studied this question.
Synapse has enriched 3 closely related papers on similar clinical questions. Consider them for comparative context: