Standard cell design system, standard cell design optimization method thereof, and semiconductor design system

Patent metadata

Field	Value
Publication number	US-10796068-B2
Application number	US-201916390087-A
Country	US
Kind code	B2
Filing date	Apr 22, 2019
Priority date	Sep 11, 2018
Publication date	Oct 6, 2020
Grant date	Oct 6, 2020

Abstract

A standard cell design system is provided. The standard cell design system includes at least one processor configured to implement: a control engine that determines planar parameters and vertical parameters of a target standard cell, a three-dimensional structure generating engine that generates a three-dimensional structure of the target standard cell based on the planar parameters and the vertical parameters, an extraction engine that extracts a standard cell model of the target standard cell from the three-dimensional structure, an assessment engine that performs a plurality of assessment operations based on the standard cell model, and an auto-optimizing engine that adjusts, based on a machine learning algorithm, the planar parameters and the vertical parameters based on results of the plurality of assessment operations.

First claim

What is claimed is:

1. A standard cell design system comprising:
at least one processor configured to implement:
a control engine configured to determine planar parameters and vertical parameters of elements provided in a target standard cell;
a three-dimensional structure generating engine configured to generate a three-dimensional structure of the target standard cell based on the planar parameters and the vertical parameters;
an extraction engine configured to extract a standard cell model of the target standard cell from the three-dimensional structure;
an assessment engine configured to perform a plurality of assessment operations based on the standard cell model; and
an auto-optimizing engine configured to adjust, based on a machine learning algorithm, the planar parameters and the vertical parameters based on results of the plurality of assessment operations,

wherein the planar parameters include planar layout information of the target standard cell, and
wherein the vertical parameters include a plurality of process information of the target standard cell.

2. The standard cell design system of claim 1 , wherein the three-dimensional structure generating engine is further configured to generate the three-dimensional structure by applying a litho contour which is based on an optical proximity correction model, and
wherein the optical proximity correction model is determined based on the planar parameters and the vertical parameters.

3. The standard cell design system of claim 1 , wherein the standard cell model includes a compact model and a parasitic extraction model associated with the target standard cell extracted from the three-dimensional structure.

4. The standard cell design system of claim 3 , wherein the parasitic extraction model is one from among a two-dimensional parasitic extraction model and a three-dimensional parasitic extraction model.

5. The standard cell design system of claim 1 , wherein the plurality of assessment operations include at least one among a ground rule assessment operation for the target standard cell, a performance-power assessment operation for the target standard cell, and a yield assessment operation for the target standard cell.

6. The standard cell design system of claim 5 , wherein the assessment engine is further configured to selectively perform some of the plurality of assessment operations based on a type of the target standard cell.

7. The standard cell design system of claim 5 , wherein the results of the plurality of assessment operations include information concerning at least one among an area of the target standard cell, a performance of the target standard cell, a power of the target standard cell, and a yield of the target standard cell, and
wherein the auto-optimizing engine is further configured to adjust the planar parameters and the vertical parameters based on the machine learning algorithm such to decrease the area of the target standard cell, improve the performance of the target standard cell, decrease the power of the target standard cell, and increase the yield of the target standard cell.

8. The standard cell design system of claim 1 , wherein the at least one processor is further configured to generate a standard cell library based on the planar parameters and the vertical parameters adjusted by the auto-optimizing engine.

9. A standard cell design optimization method of a standard cell design system, comprising:
determining planar parameters and vertical parameters of elements provided in a target standard cell;
generating a first three-dimensional structure of the target standard cell based on the planar parameters and the vertical parameters;
extracting a first standard cell model from the first three-dimensional structure;
performing a plurality of assessment operations on the target standard cell based on the first standard cell model;
determining whether results of the plurality of assessment operations satisfy a plurality of reference values, respectively;
identifying, based on the results of the plurality of assessment operations not satisfying the plurality of reference values, readjusted planar parameters and readjusted vertical parameters based on a training model updated according to the planar parameters, the vertical parameters, and the results of the plurality of assessment operations;
generating a second three-dimensional structure based on the readjusted planar parameters and the readjusted vertical parameters;
extracting a second standard cell model based on the second three-dimensional structure; and
performing the plurality of assessment operations based on the second standard cell model,
wherein the planar parameters include planar layout information of the target standard cell, and
wherein the vertical parameters include a plurality of process information of the target standard cell.

10. The method of claim 9 , wherein the first three-dimensional structure is generated by applying a litho contour which is based on an optical proximity correction model, and
wherein the optical proximity correction model is determined based on the planar parameters and the vertical parameters.

11. The method of claim 9 , wherein the plurality of assessment operations include at least one among a ground rule assessment operation for the target standard cell, a performance-power assessment operation for the target standard cell, and a yield assessment operation for the target standard cell.

12. The method of claim 11 , wherein the results of the plurality of assessment operations include information about at least one among an area of the target standard cell, a performance of the target standard cell, a power of the target standard cell, and a yield of the target standard cell,
wherein the plurality of reference values include at least one among a first reference value associated with the area of the target standard cell, a second reference value associated with the performance of the target standard cell, a third reference value associated with the power of the target standard cell, and a fourth reference value associated with the yield of the target standard cell, and
wherein the readjusted planar parameters and the readjusted vertical parameters are identified based on the training model such that the area of the target standard cell is smaller than the first reference value, the performance of the target standard cell is greater than the second reference value, the power of the target standard cell is less than the third reference value, and the yield of the target standard cell is greater than the fourth reference value.

13. The method of claim 11 , further comprising generating the training model is generated based on a result of the plurality of assessment operations which are based on planar parameters randomly sampled with regard to the target standard cell and vertical parameters randomly sampled with regard to the target standard cell.

14. A semiconductor design system comprising:
at least one processor configured to implement:
a standard cell design system configured to optimize planar parameters and vertical parameters of elements provided in a plurality of standard cells based on a first machine learning algorithm;
a standard cell library configured to generate optimized standard cell information respectively corresponding to each of the plurality of standard cells based on the planar parameters and the vertical parameters optimized by the standard cell design system; and
a block design system configured to generate a block layout of a target semiconductor device based on the optimized standard cell information generated by

Assignees

• Samsung Electronics Co Ltd

Inventors

• Kwon Uihui
• Qi Weiyi
• Lu Yang
• Ahn Saetbyeol
• Okagaki Takeshi

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06N3/08 Primary

  Learning methods · CPC title

• G06N3/092

  Reinforcement learning · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title