# US 9,519,735 B2 · Method of failure analysis

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 9,519,735 B2
- Document ID: US9519735B2
- Status: Granted
- Kind: B2
- Application number: 14/492,866
- Filed: 2014-09-22
- Published: 2015-03-26
- Granted: 2016-12-13
- Earliest claimed: 2013-09-23
- Assignees: Taiwan Semiconductor Manufacturing Company, Ltd.
- Applicants: Taiwan Semiconductor Manufacturing Co., Ltd.
- Inventors: Chin-Cheng Kuo; Kmin Hsu; Wei-Yi Hu; Wei Min Chan; Jui-Feng Kuan
- Web page: https://patentgrep.com/patent/US9519735B2

## CPC

- G06F 30/3308: using simulation
- G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30)
- G06F 30/367: Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

In some methods, a number of input data sets is provided for an integrated circuit (IC) model. A number of scores for the number of input data sets, respectively, are then determined based on probabilities of the respective input data sets resulting in a failure condition, which exists when the IC model fails to meet a predetermined yield criteria. A simulation order for the number of input data sets is then assigned according to the determined number of scores.

## Claims

### Claim 1

A simulation method, comprising:
providing a plurality of input data sets for an integrated circuit (IC) model;
performing a first simulation using a computing unit by applying a first of the plurality of input data sets to the IC model to generate a first simulation result, wherein the first input data set includes a first input device parameter with a first baseline parametric value and a second input device parameter with a second baseline parametric value;
performing a second simulation by applying a second of the plurality of input data sets to the IC model to generate a second simulation result, wherein the second input data set includes the first input device parameter but with a first parametric value that differs from the first baseline parametric value and also includes the second input device parameter with the second baseline parametric value;
determining a sensitivity of the IC model to the first input device parameter by comparing results of the first and second simulations;
defining a simulation order for remaining input data sets of the plurality of input data sets according to the sensitivity of the IC model to the first input device parameter; and
manufacturing an integrated circuit on a semiconductor substrate, wherein the manufactured integrated circuit corresponds to the IC model.

### Claim 2 (depends on 1)

The method of claim 1 further comprising:
performing a third simulation by applying a third of the plurality of input data sets to the IC model to generate a third simulation result, wherein the third input data set includes the first input device parameter with the first baseline parametric value and also includes the second input device parameter but with a second parametric value that differs from the second baseline parametric value;
determining a sensitivity of the IC model to the second input device parameter by comparing results of the first and third simulations; and
defining a simulation order for the remaining input data according to the sensitivity of the IC model to the second input device parameter.

### Claim 3 (depends on 2)

The method of claim 2:
wherein determining the sensitivity of the IC model to the first input device parameter comprises determining a first difference between the first simulation result and the second simulation result;
wherein determining the sensitivity of the IC model to the second input device parameter comprises determining a second difference between the first simulation result and the third simulation result.

### Claim 4 (depends on 3)

The method of claim 3, further comprising:
partitioning the remaining input data sets into a first subset and a second subset, wherein the input data sets of the first subset exhibit variations in the first input device parameter and the input data sets of the second subset exhibit variations in the second input device parameter; and
based on whether the IC model exhibits a higher sensitivity to the first or second input device parameter, selectively defining the simulation order for the remaining input data sets so one of the first subset or second subset is simulated prior to the other of the first subset or second subset.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
determining a first set probability of the first input data set by calculating a product of parametric value probabilities for respective input data parameters in the first input data set;
multiplying parametric value probabilities for respective input data parameters in the second input data set to determine a second set probability of the second input data set; and
defining the simulation order according to the first and second set probabilities.

### Claim 6 (depends on 5)

The method of claim 5, further comprising:
determining a total probability of each input data set as a product of the probabilities of each parametric value of each input device parameter in the input data set; and
determining the simulation order for the plurality of input data sets from the total probability of each input data set.

### Claim 7 (depends on 6)

The method of claim 6, wherein defining the simulation order for the plurality of input data sets further comprises:
summing a product of each random value of a respective input data set and its weighted sensitivity;
defining a normalization value of the set of as a square of the sum; and
defining a score of the set as the total probability of the set divided by the normalization value of the input data set.

### Claim 8 (depends on 7)

The method of claim 7, wherein the simulation order for the plurality of input data sets is determined from a lowest score to a highest score.

### Claim 9 (depends on 1)

The method of claim 1 further comprising:
based on the determined sensitivities, calculating respective scores for the remaining respective input data sets prior to simulating the respective input data sets; and
setting the simulation order so successive scores for the remaining respective input data sets follow a predetermined pattern.

### Claim 10 (depends on 9)

The method of claim 9, wherein the predetermined pattern is followed when successive scores are continuously increasing or continuously decreasing over for successive simulations for the remaining input data sets.

### Claim 11

A simulation method, comprising:
providing a plurality of input data sets for an integrated circuit (IC) model;
determining a plurality of scores for the plurality of input data sets, respectively, wherein each score is based on a probability that a corresponding input data set on which the score is based is expected to result in a failure condition where the IC model fails to meet a predetermined yield criteria;
assigning a simulation order for the plurality of input data sets according to the determined plurality of scores; and
manufacturing an integrated circuit on a semiconductor substrate, wherein the manufactured integrated circuit corresponds to the IC model.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
simulating at least some of the plurality of input data sets according to the determined plurality of scores; and
after an input data set is simulated, evaluating whether a result of that simulation meets the predetermined yield criteria.

### Claim 13 (depends on 12)

The method of claim 12, wherein if the result does not meet the predetermined criteria, halting simulations for the remaining input data sets so that revisions to the IC model can be made.

### Claim 14 (depends on 12)

The method of claim 12, wherein if the result does meet the predetermined criteria, determining if additional simulations for the input data sets remain to be carried out.

### Claim 15

A tool arrangement for circuit design, comprising:
a computing unit to process computer executable instructions; and
a memory operably coupled to the computing unit, the memory configured to store: a plurality of input data sets for an integrated circuit (IC) model, simulation ordering instructions to determine a simulation order for the plurality of input data sets based on probabilities of the respective input data sets resulting in a failure condition where the IC model fails to meet a predetermined yield criteria, and simulation instructions to carry out simulations of the input data sets in the simulation order, wherein the simulation ordering instructions are configured to assign scores to respective input data sets based on the probabilities of the respective input data sets; and
a semiconductor substrate on which the IC model is fabricated.

### Claim 16 (depends on 15)

The tool arrangement of claim 15, wherein the plurality of input data sets include randomly selected parametric values that are chosen to fall within process corners provided by a fabrication facility.

### Claim 17

A simulation method, comprising:
providing a plurality of input data sets for an integrated circuit (IC) model;
performing a first simulation by applying a first of the plurality of input data sets to the IC model to generate a first simulation result, wherein the first input data set includes a first input device parameter with a first baseline parametric value and a second input device parameter with a second baseline parametric value;
performing a second simulation by applying a second of the plurality of input data sets to the IC model to generate a second simulation result, wherein the second input data set includes the first input device parameter but with a first parametric value that differs from the first baseline parametric value and also includes the second input device parameter with the second baseline parametric value;
performing a third simulation by applying a third of the plurality of input data sets to the IC model to generate a third simulation result, wherein the third input data set includes the first input device parameter with the first baseline parametric value and also includes the second input device parameter but with a second parametric value that differs from the second baseline parametric value;
determining first and second sensitivities of the IC model to the first and second input device parameters, respectively, by comparing results of the first simulation to the second and third simulations, respectively;
calculating first and second weighted sensitivities for the first and second input data parameters, respectively, by dividing the first and second sensitivities by a larger of the first and second sensitivities;
defining a simulation order for remaining input data sets of the plurality of input data sets according to the weighted sensitivity of the IC model to the first input device parameter; and
manufacturing an integrated circuit on a semiconductor substrate, wherein the manufactured integrated circuit corresponds to the IC model.