# US 10,769,553 B2 · Integrated circuit device and circuitry

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

- Number: US 10,769,553 B2
- Document ID: US10769553B2
- Status: Granted
- Kind: B2
- Application number: 16/212,012
- Filed: 2018-12-06
- Published: 2020-06-11
- Granted: 2020-09-08
- Earliest claimed: 2018-12-06
- Assignees: Nanya Technology Corporation
- Applicants: Nanya Technology Corporation
- Inventors: Chung-Peng Hao
- Web page: https://patentgrep.com/patent/US10769553B2

## CPC

- G06N 20/00: Machine learning
- G01R 19/10: Measuring sum, difference or ratio
- G06F 1/10: Distribution of clock signals {, e.g. skew}
- G06N 3/0499: Feedforward networks
- G06N 3/065: Analogue means
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 5/04: Inference or reasoning models
- H03K 5/135: by the use of time reference signals, e.g. clock signals
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H03K: PULSE TECHNIQUE (measuring pulse characteristics G01R; modulating sinusoidal oscillations with pulses H03C; transmission of digital information H04L; discriminator circuits detecting phase difference between two signals by counting or integrating cycles of oscillation H03D3/04; automatic control, starting, synchronisation or stabilisation of generators of electronic oscillations or pulses where the type of generator is irrelevant or unspecified H03L; coding, decoding or code conversion, in general H03M)

## Abstract

The present disclosure provides an integrated circuit (IC) device and a circuitry. The IC includes a measurement circuit and a classifier circuit. The measurement circuit is configured to acquire a practical voltage. The classifier circuit is configured to: generate an information on an immature classification by comparing a default voltage and the practical voltage; receive an information on a reference classification, wherein the reference classification is acquired by manually comparing the default voltage and the practical voltage; update the default voltage to a learned voltage based on the immature classification and the reference classification; and generate a prediction, based on the learned voltage, for adjusting a slew rate.

## Claims

### Claim 1

An integrated circuit (IC) device, comprising:
a measurement circuit configured to acquire a practical voltage; and
a classifier circuit configured to:
generate an information on an immature classification by comparing a default voltage and the practical voltage;
receive an information on a reference classification, which is acquired by manually comparing the default voltage and the practical voltage;
update the default voltage to a learned voltage based on the immature classification and the reference classification; and
generate a prediction, based on the learned voltage, for adjusting a slew rate,
wherein the measurement circuit includes:
a first sample-and-hold circuit configured to sample a first voltage of a signal; and
a second sample-and-hold circuit configured to sample a second voltage of the signal, and
wherein the practical voltage is associated with the first voltage and the second voltage.

### Claim 2 (depends on 1)

The IC device of claim 1, wherein a differential of the practical voltage determines a practical slew rate of a signal, and a differential of the default voltage determines a default slew rate of the signal.

### Claim 3 (depends on 1)

The IC device of claim 1, wherein the learned voltage includes a weighted value of a neural network.

### Claim 4 (depends on 1)

The IC device of claim 1, wherein the measurement circuit further includes:
a subtractor circuit configured to provide the practical voltage by subtracting the first voltage from the second voltage.

### Claim 5 (depends on 1)

The IC device of claim 1, wherein a difference between the immature classification and the reference classification is associated with a weight of a neural network, and the learned voltage includes a weighted value of the neural network.

### Claim 6 (depends on 5)

The IC device of claim 5, wherein the classifier circuit further comprises:
a multiplier configured to generate the learned voltage by multiplying the default voltage by an updated weight, wherein the updated weight is associated with a default weight.

### Claim 7 (depends on 6)

The IC device of claim 6, wherein the updated weight is associated with an algebraic relationship between the difference and the default weight.

### Claim 8

An integrated circuit (IC) device, comprising:
a comparator having an input coupled to a practical voltage;
a subtractor circuit having an input coupled to an output of the comparator and another input coupled to a reference voltage;
a voltage divider configured to receive a voltage from the subtractor circuit; and
an adder circuit having an input coupled to an output of the voltage divider, and another input coupled to a default weight,
wherein another input of the comparator is coupled from a default voltage to a learned voltage in response to presence of a difference between the reference voltage and a voltage at an output of the subtractor circuit.

### Claim 9 (depends on 8)

The IC device of claim 8, wherein the reference voltage reflects a reference classification, which is acquired by manually comparing the default voltage and the practical voltage.

### Claim 10 (depends on 8)

The IC device of claim 8, further comprising:
an inverter having an input coupled to an output of the adder circuit; and
a multiplier circuit having an input coupled to an output of the inverter, and another input coupled to the default voltage.

### Claim 11

A circuitry, comprising:
a driver configured to drive a signal; and
a machine-learning circuit including:
a comparator having an input, wherein the input is coupled to a learned voltage in response to a discrepancy between an immature classification and a reference classification,
wherein the machine-learning circuit is configured to adjust a slew rate of the signal based on the learned voltage,
wherein the comparator is configured to generate an information on the immature classification by comparing a default voltage to a practical voltage, and
wherein the machine-learning circuit further includes:
an adder circuit configured to provide an updated weight by adding a variation reflecting the discrepancy to a default weight.

### Claim 12 (depends on 11)

The circuitry of claim 11, wherein a differential of the default voltage determines a default slew rate of the signal.

### Claim 13 (depends on 11)

The circuitry of claim 11, wherein the machine-learning circuit further includes:
a measurement circuit configured to provide the practical voltage by practically measuring the signal.

### Claim 14 (depends on 13)

The circuitry of claim 13, wherein the measurement circuit includes:
a first sample-and-hold circuit configured to sample a first voltage of the signal;
a second sample-and-hold circuit configured to sample a second voltage of the signal; and
a subtractor circuit configured to provide the practical voltage by subtracting the first voltage from the second voltage.

### Claim 15 (depends on 13)

The circuitry of claim 13, wherein a differential of the practical voltage determines a practical slew rate of the signal.

### Claim 16 (depends on 11)

The circuitry of claim 11, wherein the learned voltage is a function of the updated weight.