Multi-level hybrid algorithm filtering type branch prediction method and prediction system

Abstract

The invention discloses a multi-level hybrid algorithm filtering type branch prediction method and system, and the method comprises the steps: S1, setting two levels of branch target prediction tables, namely a first-level branch target prediction table and a second-level branch target prediction table; s2, any branch instruction accesses the first-level branch target prediction table, whether the first-level branch target prediction table is hit or not is judged, and if the first-level branch target prediction table is hit, S3 is executed; if the first-level branch target prediction table is not hit, entering S4; s3, judging whether the current predicted branch instruction is filtered out by the first-stage branch target prediction table or not, and if the current predicted branch instruction is filtered out, entering S4; otherwise, taking the prediction result provided by the first-stage branch target prediction table as the prediction address of the corresponding branch instruction; and S4, accessing the second-level branch target prediction table, and taking a prediction result provided by the second-level branch target prediction table as a prediction address of the corresponding branch instruction.

Description

technical field [0001] The present invention relates to the technical field of computer systems, in particular, to a multi-level hybrid algorithm filtering branch prediction method and prediction system. Background technique [0002] Modern processors generally use branch prediction technology to predict the direction and target address of the branch before the branch instruction is executed. If the prediction is a jump, the instruction is fetched in advance according to the predicted address to avoid pipeline bubbles when the branch occurs and affect performance. The principle of branch prediction is to save the historical information of the branch instruction. When the same branch instruction is encountered again, the behavior of the branch instruction is predicted according to the saved historical information. [0003] The impact of branch instructions on performance mainly comes from two parts: the first part comes from the time consumed by branch prediction, that is, th...

AI Technical Summary

Problems solved by technology

but figure 1 The prediction method shown does not distinguish the difficulty of branch prediction, so the two prediction tables in the two-level predictor need to predict all branches at the same time, so the power consumption of the predictor still cannot be reduced

Images