# US 2017/0255732 A1 · Memory and Logic Lifetime Simulation Systems and Methods

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

- Number: US 2017/0255732 A1
- Document ID: US20170255732A1
- Status: Published application
- Kind: A1
- Application number: 15/446,799
- Filed: 2017-03-01
- Published: 2017-09-07
- Earliest claimed: 2016-03-01
- Assignees: Georgia Tech Research Corporation
- Applicants: Georgia Tech Research Corporation
- Inventors: Linda Milor; Taizhi Liu; Chang-Chih Chen
- Web page: https://patentgrep.com/patent/US20170255732A1

## CPC

- G06F 11/079: {Root cause analysis, i.e. error or fault diagnosis (in a hardware test environment G06F11/22; in a software test environment G06F11/36)}
- G06F 11/008: {Reliability or availability analysis}
- G06F 11/073: {in a memory management context, e.g. virtual memory or cache management (memory management G06F12/00; testing of static memory units G11C29/00)}
- G06F 30/33: Design verification, e.g. functional simulation or model checking
- G06F 30/367: Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
- G11C 29/02: Detection or location of defective auxiliary circuits, e.g. defective refresh counters
- G11C 29/12005: {comprising voltage or current generators}
- G11C 29/24: Accessing extra cells, e.g. dummy cells or redundant cells
- G11C 29/50004: {of threshold voltage}
- G11C 29/50016: {of retention}
- G11C 29/70: {Masking faults in memories by using spares or by reconfiguring}
- G11C 2029/1208: {Error catch memory}
- G11C 2029/3602: {Pattern generator}
- G11C 2029/5002: {Characteristic}
- G11C 2029/5004: {Voltage}
- G11C 2029/5006: {Current}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G11C: STATIC STORES (semiconductor memory devices H10B)

## Abstract

Aspects of the disclosed technology include a method including extracting, by a processor, a plurality of features from one from among a layout of a circuit, a netlist of the circuit, and the layout and the netlist of the circuit; computing, by the processor, respective lifetime distributions of the plurality of extracted features based on at least one circuit profile; and estimating, by the processor, a lifetime of the circuit by combining the respective lifetime distributions of the plurality of extracted features.

## Claims

### Claim 1

A lifetime estimation method comprising:
extracting, by a processor, a plurality of features from one from among a layout of a circuit, a netlist of the circuit, or the layout and the netlist of the circuit;
computing, by the processor, respective lifetime distributions of the plurality of extracted features based on at least one circuit profile; and
estimating, by the processor, a lifetime of the circuit by combining the respective lifetime distributions of the plurality of extracted features.

### Claim 2 (depends on 1)

The method of claim 1, wherein the at least one circuit profile comprises one or more from among a thermal profile, an activity profile, a state profile, a voltage source profile, and an ohmic potential drop (IR drop) profile.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
counting, by the processor, a number of the plurality of features in the circuit layout; and
identifying, by the processor, a location of the plurality of features,
wherein the estimating is based, at least in part, on the number and location of the plurality of features.

### Claim 4 (depends on 1)

The method of claim 1, wherein
the circuit comprises memory, and
the computing respective lifetime distributions of the plurality of extracted features varies, at least in part, as a function of at least one from among cache associatively, cache line size, cache size, and cache replacement algorithm.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
modeling, by the processor, at least one block of the circuit using multivariate adaptive regression splines (MARS) as a fitted function between a response, process parameters, environmental parameters, and timing parameters,
wherein the computing the respective lifetime distributions is based, at least in part, on at least one of the at least modeled one circuit block.

### Claim 6 (depends on 1)

The method of claim 1, wherein the estimating the lifetime of the circuit comprises estimating the effect of error correction on the circuit.

### Claim 7 (depends on 2)

The method of claim 2, further comprising:
generating, based on the activity profile, the stress profile, and the thermal profile, respective bias temperature instability (BTI) profiles of the plurality of features; and
generating, based on the activity profile and the stress profile, and the thermal profile, respective hot carrier injection (HCI) profiles of the plurality of features,
wherein the computing the respective lifetime distributions is based, at least in part, on the combined effect of the BTI profiles and the HCI profiles.

### Claim 8 (depends on 2)

The method of claim 2, further comprising calculating, by the processor using statistical simulation, a time to failure distribution of a memory cell of the circuit based on at least one from among the thermal profile, the activity profile, the state profile, the voltage source profile, the IR drop profile, and at least one process parameter.

### Claim 9 (depends on 2)

The method of claim 2, further comprising:
building, by the processor, a gate-delay model using multivariate adaptive regression splines (MARS) as a fitted function between response variables and process-voltage-temperature (PVT) variations across the circuit,
wherein the computing the respective lifetime distributions is based, at least in part, on the gate-delay model.

### Claim 10 (depends on 8)

The method of claim 8, wherein the time to failure distribution of the memory is calculated for a plurality of discrete values within the thermal profile, the activity profile, the state profile, the voltage source profile, the IR drop profile, and at least one of the at least one process parameter.

### Claim 11 (depends on 9)

The method of claim 9, further comprising:
modeling, by the processor, gates and interconnects of the circuit using MARS and a variational Pi-model load,
wherein the computing the respective lifetime distributions is further based, at least in part, on the gate and interconnect models.

### Claim 12 (depends on 9)

The method of claim 9, wherein the gate delay model further uses MARS as a fitted function between response variables and the PVT variations across the circuit manufacturing process.

### Claim 13 (depends on 9)

The method of claim 9, further comprising:
extracting, by the processor using block-based statistical timing analysis (StTA), critical paths; and
generating, by the processor based on the critical paths, at least one circuit component-delay distribution,
wherein the estimating is based, at least in part, on the at least one circuit component-delay distribution.

### Claim 14 (depends on 13)

The method of claim 13, wherein the estimating the lifetime of the circuit is based, at least in part, on a combination of the at least one circuit component-delay distribution with a timing requirement of the circuit.

### Claim 15

A method comprising:
simulating, by a processor, a plurality of accelerated aging tests of a circuit over respective ranges of a plurality of parameters;
determining, by the processor, one or more wearout mechanisms of the circuit that can cause failures for each accelerated aging test; and
identifying, by the processor, a selectivity region defined by sub-ranges of the plurality of parameters for which a specific wearout mechanism of the plurality of wearout mechanisms is most likely to occur.

### Claim 16 (depends on 15)

The method of claim 15, further comprising:
selecting testing parameters for an accelerated aging test from within the selectivity region; and
performing the accelerated aging test on a physical circuit.

### Claim 17 (depends on 15)

The method of claim 15, wherein
the plurality of parameters comprise temperature and operating voltage, and
the specific wearout mechanism comprises one from among gate oxide breakdown (GTDDB), backend dielectric breakdown (BTDDB), and middle-of-line time-dependent dielectric breakdown (MTDDB).

### Claim 18

A lifetime estimation method comprising:
receiving, by a processor, a netlist of a circuit;
identifying, by the processor and based on the received netlist, a plurality of features of the circuit that are vulnerable to a plurality of wearout mechanisms;
computing, by the processor, respective lifetime distributions of the plurality of extracted features based on at least one circuit profile, the respective lifetime distributions being based, at least in part, on a combined effect of the plurality of wearout mechanisms; and
estimating, by the processor, a lifetime of the circuit by combining the respective lifetime distributions of the plurality of extracted features.

### Claim 19 (depends on 18)

The method of claim 18, wherein the computing the respective lifetime distributions comprises:
simulating, by the processor, a plurality of accelerated tests on the circuit; and
identifying, by the processor, respective failures of the plurality of simulated accelerated tests.

### Claim 20

A lifetime estimation method comprising:
receiving, by a processor, a netlist of a circuit;
generating, by the processor and based on at least one first circuit profile, a bias temperature instability (BTI) profile of the circuit;
generating, by the processor and based on at least one second circuit profile, a hot carrier injection (HCI) profile of the circuit;
simulating, by the processor, a plurality of tests on the circuit, the simulating being based, at least in part, on a combined effect of the BTI profiles and the HCI profiles; and
estimating, by the processor, a lifetime of the circuit, based on the simulated plurality of tests.

### Claim 21 (depends on 20)

The method of claim 20, wherein
the simulating comprises simulating, by the processor, the plurality of tests through the circuit failing to satisfy a defined performance metric, and
the defined performance metric comprises at least one from among a write margin, read and retention static noise margins (SNMs), a write margin, a read current, a minimum retention voltage, and circuit delay.
