The performance potential of value and dependence prediction

Abstract

The serialization constraints induced by the detection and enforcement of true data dependences have always been regarded as requirements for correct execution. We propose two data-speculative techniques-source operand value prediction and dependence prediction-that can be used to relax these constraints to allow instructions to execute before their data dependences are resolved or even detected. We find that inter-instruction dependences and source operand values are easily predictable. These discoveries minimize the per-cycle instruction throughput (or IPC) penalty of deeper pipelining of instruction dispatch and result in average integer program speedups ranging from 22% to 106%, depending on machine issue width and pipeline depth.