Dynamic data dependence tracking and its application to branch prediction

To continue to improve processor performance, microar-chitects seek to increase the effective instruction level paral-lelism (ILP) that can be exploited in applications. A funda-mental limit to improving ILP is data dependences among instructions. If data dependence information is available at run-time, there are many uses to improve ILP. Prior pub-lished examples include decoupled branch execution archi-tectures and critical instruction detection. In this paper, we describe an efficient hardware mecha-nism to dynamically track the data dependence chains of the instructions in the pipeline. This information is available on a cycle-by-cycle basis to the microengine for optimizing its performance. We then use this design in a new value-based branch prediction design using Available Register Value In-formation (ARVI). From the use of data dependence infor-mation, the ARVI branch predictor has better prediction accuracy over a comparably sized hybrid branch predic-tor. With ARVI used as the second-level branch predictor, the improved prediction accuracy results in a 12.6 % perfor-mance improvement on average across the SPEC95 integer benchmark suite.