A time sequence path correction method

Abstract

The invention discloses a time sequence path correction method. The method comprises the following steps: determining a path and a unit of time sequence violation; setting a search range and searching an exchangeable unit; exchanging the time sequence violation units and the exchangeable units, evaluating new time sequence values of the two time sequence paths, and adding the exchangeable units which do not violate the time sequence and weight values of the exchangeable units into a result set; and in the result set, selecting the exchangeable unit with the maximum weight value to be exchanged with the time sequence violation unit to generate a result path. According to the time sequence path correction method, on the premise that the chip design physical layout is not changed and the chip function behavior is not changed, the purpose of time sequence optimization is achieved through unit exchange, so that time sequence correction can be carried out in the post-mask stage in the design later stage, and the correctness of chip design is guaranteed.

Description

technical field [0001] The present invention relates to the technical field of electronic design automation (EDA), in particular to a method for modifying timing without changing layout unit swap. Background technique [0002] In digital integrated circuit design, in order to ensure that the chip can work normally and achieve the expected frequency, it is necessary to check whether the time when the clock signal and data signal arrive at the register synchronization unit meets the constraints of setup time and hold time. If timing violations are found, ECO modifications are required to adjust the timing paths. [0003] Buffer cell insertion, cell size change, and large net splitting are commonly used timing optimization methods. For a timing path with a timing violation, the delay of the timing path is increased or decreased by means of cell insertion or cell size change, so as to correct the timing violation problem. However, the premise of this timing optimization is tha...

AI Technical Summary

Problems solved by technology

However, when the number of spare spare units is insufficient, or the spare spare unit cannot be found within the specified range, the timing target cannot be optimized

Images