ADAPTIVE GRANULARITY - TRANSPARENT INTEGRATION OF FINE-GRAIN AND COARSE-GRAIN COMMUNICATION

The granularity of shared data is one of the key Factors affecting the performance of distributed shared memory machines (DSM). Given that p rograms exhibit quite different sharing patterns, providing only one o r two fixed granularities cannot result in an efficient use of resourc es. On the other hand, supporting arbitrarily granularity sizes signif icantly increases not only hardware complexity but software overhead a s well. Furthermore. the efficient use of arbitrarily granularities pu t the burden on users to provide information about program behavior to compilers and/or runtime systems. These kind of requirements tend to restrict the programmability of the shared memory model. In this paper , we present a new communication scheme, called Adaptive Granularity ( AG). Adaptive Granularity makes it possible to transparently integrate bulk transfer into the shared memory model by supporting variable-siz e granularity and memory replication. It consists of two protocols: on e for small data and another for large data. For small size data, the standard hardware DSM protocol is used and the granularity is fixed to the size of a cache line. For large array data, the protocol for bulk data is used instead, and the granularity varies depending on the run time sharing behavior of the applications. Simulation results show tha t AG improves performance up to 43% over the hardware implementation o f DSM (e.g., DASH, Alewife). Compared with an equivalent architecture that supports fine-grain memory replication at the fixed granularity o f a cache line (e.g., Typhoon), AG reduces execution time up to 35%.