Server-Based Data Push Architecture for Multi-Processor Environments

Data access delay is a major bottleneck in utilizing currenthigh-end computing (HEC) machines. Prefetching, where data is fetchedbefore CPU demands for it, has been considered as an effective solutionto masking data access delay. However, current client-initiatedprefetching strategies, where a computing processor initiatesprefetching instructions, have many limitations. They do not work wellfor applications with complex, non-contiguous data access patterns.While technology advances continue to increase the gap betweencomputing and data access performance, trading computing power forreducing data access delay has become a natural choice. In this paper,we present a server-based data-push approach and discuss its associatedimplementation mechanisms. In the server-push architecture, a dedicatedserver called Data Push Server (DPS) initiates and proactively pushesdata closer to the client in time. Issues, such as what data to fetch,when to fetch, and how to push are studied. The SimpleScalar simulatoris modified with a dedicated prefetching engine that pushes data foranother processor to test DPS based prefetching. Simulation resultsshow that L1 Cache miss rate can be reduced by up to 97\% (71\% onaverage) over a superscalar processor for SPEC CPU2000 benchmarks thathave high cache miss rates.