Exploiting ILP in page-based intelligent memory

Abstract

This study compares the speed, area, and power of di#erent implementations of ActivePages #OCS98#, an intelligent memory system which helps bridge the growing gap between processor and memory performance by associating simple functions with each page of data. Previous investigations have shown up to 1000X speedups using a block of recon#gurable logic to implement these functions next to each subarray on a DRAM chip. In this study,we show that instruction-level parallelism, not hardware specialization, is the key to the previous success with recon#gurable logic. In order to demonstrate this fact, an ActivePage implementation based upon a simpli#ed VLIW processor was developed. Unlike conventional VLIW processors, power and area constraints lead to a design which has a small number of pipeline stages. Our results demonstrate that a four-wide VLIW processor attains comparable performance to that of pure FPGA logic but requires signi#- cantly less area and power. 1