# US 2021/0344571 A1 · Method, device, and computer program product for processing data

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

- Number: US 2021/0344571 A1
- Document ID: US20210344571A1
- Status: Published application
- Kind: A1
- Application number: 16/883,088
- Filed: 2020-05-26
- Published: 2021-11-04
- Earliest claimed: 2020-04-30
- Assignees: EMC IP Holding Company LLC
- Applicants: EMC IP Holding Company LLC
- Inventors: Chenxi Hu; Sanping Li
- Web page: https://patentgrep.com/patent/US20210344571A1

## CPC

- H04L 41/16: using machine learning or artificial intelligence
- H04L 43/0876: Network utilisation, e.g. volume of load or congestion level
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- H04L 41/145: {involving simulating, designing, planning or modelling of a network}
- H04L 43/0852: Delays
- H04L 45/563: {Software download or update}
- G06N 3/08: Learning methods
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Embodiments of the present disclosure relate to a method, a device, and a computer program product for processing data. The method includes: loading, at a switch and in response to receipt of a model loading request from a terminal device, a data processing model specified in the model loading request. The method further includes: acquiring model parameters of the data processing model from the terminal device. The method further includes: processing, in response to receipt of to-be-processed data from the terminal device, the data using the data processing model based on the model parameters. Through the method, data may be processed at a switch, which improves the efficiency of data processing and the utilization rate of computing resources, and reduces the delay of data processing.

## Claims

### Claim 1

A method for processing data, comprising:
loading, at a switch and in response to receipt of a model loading request from a terminal device, a data processing model specified in the model loading request;
acquiring model parameters of the data processing model from the terminal device; and
processing, in response to receipt of to-be-processed data from the terminal device, the data using the data processing model based on the model parameters.

### Claim 2 (depends on 1)

The method according to claim 1, wherein loading the data processing model comprises:
acquiring an identifier of the data processing model in the model loading request;
selecting, based on the identifier, the data processing model from a predetermined data processing model set; and
loading the selected data processing model.

### Claim 3 (depends on 1)

The method according to claim 1, wherein acquiring the model parameters comprises:
determining, based on a content label of a packet received from the terminal device, whether the packet comprises model parameters related to the data processing model; and
acquiring the model parameters from the packet according to a determination that the packet comprises the model parameters.

### Claim 4 (depends on 3)

The method according to claim 3, wherein the data processing model is a neural network model, and acquiring the model parameters from the packet comprises:
determining a parameter size corresponding to each processing layer of the neural network model; and
acquiring model parameters corresponding to each processing layer from the packet according to the parameter size.

### Claim 5 (depends on 1)

The method according to claim 1, further comprising:
determining, based on a content label of a packet received from the terminal device, whether the packet comprises data to be processed by the data processing model; and
determining, according to a determination that the packet comprises data to be processed by the data processing model, that the to-be-processed data is received.

### Claim 6 (depends on 1)

The method according to claim 1, wherein the data processing model is a neural network model, and processing the data comprises:
processing, based on model parameters corresponding to each processing layer of the neural network model, the data using the neural network model.

### Claim 7 (depends on 1)

The method according to claim 1, further comprising at least one of the following:
sending, in response to receipt of a destination address from the terminal device, a processing result for the data based on the destination address; and
sending the processing result for the data to the terminal device.

### Claim 8

A method for processing data, comprising:
sending, at a terminal device, a model loading request to a switch, a data processing model to be loaded by the switch being specified in the model loading request;
sending model parameters of the data processing model to the switch; and
sending to-be-processed data to the switch, such that the to-be-processed data is processed at the switch based on the data processing model.

### Claim 9 (depends on 8)

The method according to claim 8, further comprising:
receiving a processing result for the data from the switch.

### Claim 10 (depends on 8)

The method according to claim 8, wherein sending the model parameters to the switch comprises:
including the model parameters in a packet to be sent;
setting a content label of the packet as indicating that the packet comprises the model parameters; and
sending the packet to the switch.

### Claim 11 (depends on 8)

The method according to claim 8, wherein sending to-be-processed data to the switch comprises:
including the to-be-processed data in a packet to be sent;
setting a content label of the packet as indicating that the packet comprises the to-be-processed data; and
sending the packet to the switch.

### Claim 12

A switch, comprising:
a processor; and
a memory storing computer program instructions, wherein the processor runs the computer program instructions in the memory to control an electronic device to perform actions comprising:
loading, in response to receipt of a model loading request from a terminal device, a data processing model specified in the model loading request;
acquiring model parameters of the data processing model from the terminal device; and
processing, in response to receipt of to-be-processed data from the terminal device, the data using the data processing model based on the model parameters.

### Claim 13 (depends on 12)

The switch according to claim 12, wherein loading the data processing model comprises:
acquiring an identifier of the data processing model in the model loading request;
selecting, based on the identifier, the data processing model from a predetermined data processing model set; and
loading the selected data processing model.

### Claim 14 (depends on 12)

The switch according to claim 12, wherein acquiring the model parameters comprises:
determining, based on a content label of a packet received from the terminal device, whether the packet comprises model parameters related to the data processing model; and
acquiring the model parameters from the packet according to a determination that the packet comprises the model parameters.

### Claim 15 (depends on 14)

The switch according to claim 14, wherein the data processing model is a neural network model, and acquiring the model parameters from the packet comprises:
determining a parameter size corresponding to each processing layer of the neural network model; and
acquiring model parameters corresponding to each processing layer from the packet according to the parameter size.

### Claim 16 (depends on 12)

The switch according to claim 12, wherein the actions further comprise:
determining, based on a content label of a packet received from the terminal device, whether the packet comprises data to be processed by the data processing model; and
determining, according to a determination that the packet comprises data to be processed by the data processing model, that the to-be-processed data is received.

### Claim 17 (depends on 12)

The switch according to claim 12, wherein the data processing model is a neural network model, and processing the data comprises:
processing, based on model parameters corresponding to each processing layer of the neural network model, the data using the neural network model.

### Claim 18 (depends on 12)

The switch according to claim 12, wherein the actions further comprise at least one of the following:
sending, in response to receipt of a destination address from the terminal device, a processing result for the data based on the destination address; and
sending the processing result for the data to the terminal device.

### Claim 19 (depends on 1)

A computer program product tangibly stored on a non-volatile computer-readable medium and comprising machine-executable instructions, wherein the machine-executable instructions, when executed, cause a machine to perform the method of claim 1.

### Claim 20 (depends on 8)

A computer program product tangibly stored on a non-volatile computer-readable medium and comprising machine-executable instructions, wherein the machine-executable instructions, when executed, cause a machine to perform the method of claim 8.