# US 2024/0176616 A1 · Processor and method of controlling processor

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

- Number: US 2024/0176616 A1
- Document ID: US20240176616A1
- Status: Published application
- Kind: A1
- Application number: 18/519,635
- Filed: 2023-11-27
- Published: 2024-05-30
- Earliest claimed: 2022-11-28
- Assignees: Preferred Networks, Inc.; National University Corporation Kobe University
- Applicants: Preferred Networks, Inc.; National University Corporation Kobe University
- Inventors: Kenji TAGATA; Harunobu MIYASHITA; Junichiro MAKINO
- Web page: https://patentgrep.com/patent/US20240176616A1

## CPC

- G06F 9/3001: {Arithmetic instructions}
- G06F 9/30032: {Movement instructions, e.g. MOVE, SHIFT, ROTATE, SHUFFLE}
- G06F 9/30043: {LOAD or STORE instructions; Clear instruction}
- G06F 9/38: Concurrent instruction execution, e.g. pipeline or look ahead
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A processor includes an arithmetic circuit configured to execute an arithmetic instruction; and a register configured to hold data used by the arithmetic circuit. The processor receives a data movement instruction and the arithmetic instruction corresponding to the data movement instruction, and moves the data from a first memory to the register based on a data movement instruction. The arithmetic circuit executes the arithmetic instruction after a data movement of the data movement instruction is completed.

## Claims

### Claim 1

A processor comprising:
an arithmetic circuit configured to execute an arithmetic instruction; and
a register configured to hold data used by the arithmetic circuit,
wherein the processor receives a data movement instruction and the arithmetic instruction corresponding to the data movement instruction, and moves the data from a first memory to the register based on a data movement instruction; and
wherein the arithmetic circuit executes the arithmetic instruction after a data movement of the data movement instruction is completed.

### Claim 2 (depends on 1)

The processor as claimed in claim 1,
wherein an instruction sequence received by the processor includes the data movement instruction, a wait instruction, and the arithmetic instruction that correspond to one another, and
wherein the processor prevents the arithmetic instruction subsequent to the wait instruction from being input to the arithmetic circuit until a corresponding completion information is input.

### Claim 3 (depends on 2)

The processor as claimed in claim 2, wherein the processor inputs the arithmetic instruction between the data movement instruction and the wait instruction to the arithmetic circuit without the preventing of the arithmetic instruction.

### Claim 4 (depends on 2)

The processor as claimed in claim 2, wherein the data movement instruction and the wait instruction include identification information for associating the data movement instruction with the wait instruction.

### Claim 5 (depends on 1)

The processor as claimed in claim 1, further comprising a second memory having a constant access latency,
wherein an access latency of the first memory is not constant, and
wherein the processor controls movement of the data for the first memory or the second memory based on the data movement instruction, and
wherein the data movement instruction is any one of an instruction to move the data between the first memory and the register, an instruction to move the data between the second memory and the register, and an instruction to move the data between the first memory and the second memory.

### Claim 6 (depends on 1)

The processor as claimed in claim 1, further comprising a network interface configured to control communication with a network,
wherein an access latency of the first memory is not constant, and
wherein the first memory is arranged in the network.

### Claim 7 (depends on 1)

The processor as claimed in claim 1, wherein the data movement instruction includes movement amount information indicating a movement amount of the data.

### Claim 8 (depends on 1)

The processor as claimed in claim 1, wherein the processor inputs permission information for permitting an input of the arithmetic instruction corresponding to the completion of the data movement of the data movement instruction based on reception of the completion information,
wherein the processor inputs the arithmetic instruction corresponding to the permission information to the arithmetic circuit based on reception of the permission information.

### Claim 9 (depends on 1)

The processor as claimed in claim 1, further comprising:
a data movement control circuit configured to move the data from a first memory to the register based on the data movement instruction; and
an instruction distribution circuit configured to receive the data movement instruction and the arithmetic instruction corresponding to the data movement instruction, output the data movement instruction, and output the arithmetic instruction to the arithmetic circuit after the data movement of the data movement instruction is completed.

### Claim 10 (depends on 9)

The processor as claimed in claim 9, wherein the completion of the data movement of the data movement instruction is informed by outputting a completion information from the data movement control circuit to the instruction distribution circuit.

### Claim 11

A method of controlling a processor including an arithmetic circuit configured to execute an arithmetic instruction and a register configured to hold data used by the arithmetic circuit, comprising:
receiving, by the processor, the data movement instruction and the arithmetic instruction corresponding to the data movement instruction, and
moving, by the processor, the data from a first memory to the register based on a data movement instruction; and
executing, by the processor, the arithmetic instruction after a data movement of the data movement instruction is completed.

### Claim 12 (depends on 11)

The method of controlling the processor as claimed in claim 11,
wherein an instruction sequence received by the processor includes the data movement instruction, a wait instruction, and the arithmetic instruction that correspond to one another, and further comprising:
preventing, by the processor, the arithmetic instruction subsequent to the wait instruction from being inputted to the arithmetic circuit until a corresponding completion information is inputted.

### Claim 13 (depends on 12)

The method of controlling the processor as claimed in claim 12, wherein the processor inputs the arithmetic instruction between the data movement instruction and the wait instruction to the arithmetic circuit without the preventing of the arithmetic instruction.

### Claim 14 (depends on 12)

The method of controlling the processor as claimed in claim 12, wherein the data movement instruction and the wait instruction include identification information for associating the data movement instruction with the wait instruction.

### Claim 15 (depends on 11)

The method of controlling the processor as claimed in claim 11, further comprising a second memory configured to have a constant access latency,
wherein an access latency of the first memory is not constant, and
wherein the processor controls movement of the data for the first memory or the second memory based on the data movement instruction, and
wherein the data movement instruction is any one of an instruction to move the data between the first memory and the register, an instruction to move the data between the second memory and the register, and an instruction to move the data between the first memory and the second memory.

### Claim 16 (depends on 11)

The method of controlling the processor as claimed in claim 11, further comprising a network interface configured to control communication with a network,
wherein an access latency of the first memory is not constant, and
wherein the first memory is arranged in the network.

### Claim 17 (depends on 11)

The method of controlling the processor as claimed in claim 11, wherein the data movement instruction includes movement amount information indicating a movement amount of the data.

### Claim 18 (depends on 11)

The method of controlling the processor as claimed in claim 11, the processor inputs permission information for permitting an input of the arithmetic instruction corresponding to the completion of a data movement of the data movement instruction based on reception of the completion information,
wherein the processor inputs the arithmetic instruction corresponding to the permission information to the arithmetic circuit based on reception of the permission information.

### Claim 19 (depends on 11)

The method of controlling the processor as claimed in claim 11, further comprising:
moving, by a data movement control circuit, the data from a first memory to the register based on the data movement instruction; and
receiving, by an instruction distribution circuit, the data movement instruction and the arithmetic instruction corresponding to the data movement instruction, outputting, by the instruction distribution circuit, the data movement instruction, and outputting, by the instruction distribution circuit, the arithmetic instruction to the arithmetic circuit after the data movement of the data movement instruction is completed.

### Claim 20 (depends on 19)

The method of controlling the processor as claimed in claim 19, wherein the completion of the data movement of the data movement instruction is informed by outputting a completion information from the data movement control circuit to the instruction distribution circuit.
