A High-Level Synthetic Scheduling Method Based on Time Constraints

Abstract

A high-level comprehensive scheduling method based on time constraints, related to integrated circuits, the present invention includes the following steps: 1) calculating the time frame of each operation; 2) determining the time frame distribution relationship of each operation and each control step; 3) for each operation Carry out classification to form N operation types; 4) form a set of control steps to be optimized with the largest degree of congestion and associated with the operation to be processed; 5) calculate the congestion degree difference of each control step in the set to be processed, 6) Adjust the currently selected operation to the control step with the least congestion of the operation type it belongs to; 7) judge whether there is an operation to be processed, if not, end; otherwise, update the congestion distribution relationship of various operations in each control step, and then Go back to step 4). The invention has the advantages of small calculation amount, remarkably improved running speed, and little difference between the obtained scheduling result and the optimal solution result.

Description

Technical field [0001] The present invention relates to an integrated circuit technology. Background technique [0002] With the continuous development of semiconductor technology, the scale of the FPGA device is growing, and the requirements for high-level synthesis are also increased. It is considered an important part of high-level integration due to scheduling issues and other tasks in high-level syndrome. The optimal solution to the operation schedule under the constraint is an NP full problem. [0003] Commonly used high-level integrated scheduling algorithms mainly include ASAP (AS Soon as Possible), ALAP (Aslate as Possible), and FDS, Force-Directed Scheduling. The ASAP and ALAP algorithms are simple and fast, but the resulting scheduling results are uneven, and the resource utilization is low. The force boot algorithm can achieve the scheduling of time constraints, which can find the best scheduling result, but the algorithm has a high complexity, and the amount is large...

AI Technical Summary

Problems solved by technology

ASAP and ALAP algorithms are simple and fast, but the operation distribution in the obtained scheduling results is uneven, and the resource utilization rate is low

The force-guided algorithm can realize scheduling under time constraints and find the best scheduling result, but the algorithm complexity is high and the amount of calculation is large

Images