Variable Block Size Architecture for Programs

Cache reconfiguration is well studied topic in recent time. This paper proposes an algorithm to determine variable block size for variables in a program at some predetermined points called decision points based on their access pattern. The whole program is divided into segments by the decision points. Rules to decide the decision points are developed. The algorithm identifies the decision points, formulates optimization function to determine the average memory access time for the variables involved at these decision points. Solving the optimization function with constraints gives the optimal block size. The algorithm is simulated for a chosen program and an improvement of 78% in AMAT is observed over direct mapped cache in SPEC2K benchmark. An improvement of 28% over prefetching was observed for the chosen program using SPEC2K benchmark.