# US 2017/0286108 A1 · Processor Core, Processor And Method For Executing A Composite Scalar-Vector Very Lare Instruction Word (VLIW) Instruction

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

- Number: US 2017/0286108 A1
- Document ID: US20170286108A1
- Status: Published application
- Kind: A1
- Application number: 15/086,947
- Filed: 2016-03-31
- Published: 2017-10-05
- Earliest claimed: 2016-03-31
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Bruce M. Fleischer; Thomas Winters FOX; Arpith C. JACOB; Hans Mikael Jacobson; Ravi Nair; Kevin John Patrick O'Brien; Daniel Arthur Prener
- Web page: https://patentgrep.com/patent/US20170286108A1

## CPC

- G06F 9/30036: {Instructions to perform operations on packed data, e.g. vector, tile or matrix operations}
- G06F 9/3853: {of compound instructions}
- G06F 9/30065: {Loop control instructions; iterative instructions, e.g. LOOP, REPEAT}
- G06F 9/3822: {Parallel decoding, e.g. parallel decode units}
- G06F 9/3836: {Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A processor core includes a storage device which stores a composite very large instruction word (VLIW) instruction, an instruction unit which obtains the composite VLIW instruction from the storage device and decodes the composite VLIW instruction to determine an operation to perform, and a composite VLIW instruction execution unit which executes the decoded composite VLIW instruction to perform the operation.

## Claims

### Claim at position 1

1: A processor core comprising:
a storage device which stores a composite very large instruction word (VLIW) instruction;
an instruction unit which obtains the composite VLIW instruction from the storage device and decodes the composite VLIW instruction to determine an operation to perform; and
a composite VLIW instruction execution unit which executes the decoded composite VLIW instruction to perform the operation.

### Claim at position 2 (depends on 1)

2: The processor core of claim 1, wherein the composite VLIW instruction comprises a scalar atom and a vector atom.

### Claim at position 3 (depends on 1)

3: The processor core of claim 1, wherein the composite VLIW instruction execution unit comprises a detector which determines whether a REPEAT field of the composite VLIW instruction is REP=1 or REP>1, and
wherein if the detector determines that the REPEAT field in the composite VLIW instruction is REP=1, then the composite VLIW instruction execution unit executes each atom in the composite VLIW instruction at a single iteration.

### Claim at position 4 (depends on 3)

4: The processor core of claim 3, wherein the composite VLIW instruction execution unit further comprises a composite VLIW instruction execution subunit, and if the detector determines that the REPEAT field in the composite VLIW instruction is REP>1, then the composite VLIW instruction execution subunit executes scalar atoms of the decoded composite VLIW instruction at a single iteration.

### Claim at position 5 (depends on 4)

5: The processor core of claim 4, wherein if the detector determines that the REPEAT field in the composite VLIW instruction is REP>1, then the composite VLIW instruction execution subunit executes vector atoms of the decoded composite VLIW instruction during all REP iterations.

### Claim at position 6 (depends on 5)

6: The processor core of claim 5, wherein if the detector determines that the REPEAT field in the composite VLIW instruction is REP>1, then the composite VLIW instruction execution subunit executes branch and control atoms of the decoded composite VLIW instruction at a single iteration.

### Claim at position 7 (depends on 6)

7: The processor core of claim 6, wherein the single iteration at which the composite VLIW instruction execution subunit executes branch and control atoms of the decoded composite VLIW instruction comprises the last iteration.

### Claim at position 8 (depends on 4)

8: The processor core of claim 4, wherein the single iteration at which the composite VLIW instruction execution subunit executes scalar atoms of the decoded composite VLIW instruction comprises the first iteration.

### Claim at position 9 (depends on 6)

9: The processor core of claim 6, wherein the branch and control atoms of the decoded composite VLIW instruction comprise a field indicating the iteration at which the composite VLIW instruction execution subunit executes the branch and control atoms.

### Claim at position 10 (depends on 4)

10: The processor core of claim 4, wherein the scalar atoms of the decoded composite VLIW instruction comprise a field indicating the iteration at which the composite VLIW instruction execution subunit executes the scalar atoms.

### Claim at position 11 (depends on 6)

11: The processor core of claim 6, wherein the composite VLIW instruction execution unit comprises an iteration setting unit which, based on a user input:
adjusts a default setting of an iteration at which to execute the branch and control atoms of the composite VLIW instruction; and
adjusts a default setting of an iteration at which to execute the scalar atoms of the composite VLIW instruction.

### Claim at position 12 (depends on 1)

12: A processor comprising the processor core of claim 1.

### Claim at position 13

13: A method of executing a composite very large instruction word (VLIW) instruction, the method comprising:
storing the composite VLIW instruction;
obtaining the composite VLIW instruction, and decoding the composite VLIW instruction to determine an operation to perform; and
executing the decoded composite VLIW instruction to perform the operation.

### Claim at position 14 (depends on 13)

14: The method of claim 13, further comprising:
determining whether a REPEAT field of the composite VLIW instruction is REP=1 or REP>1; and
if the REPEAT field in the composite VLIW instruction is determined to be REP=1, then executing each atom in the composite VLIW instruction at a single iteration.

### Claim at position 15 (depends on 14)

15: The method of claim 14, further comprising:
if the REPEAT field in the composite VLIW instruction is determined to be REP>1, then executing scalar atoms of the decoded composite VLIW instruction at a single iteration.

### Claim at position 16 (depends on 15)

16: The method of claim 15, further comprising:
if the REPEAT field in the composite VLIW instruction is determined to be REP>1, then executing vector atoms of the decoded composite VLIW instruction during all REP iterations.

### Claim at position 17 (depends on 16)

17: The method of claim 16, further comprising:
if the REPEAT field in the composite VLIW instruction is determined to be REP>1, then executing branch and control atoms of the decoded composite VLIW instruction at a single iteration.

### Claim at position 18

18: A method of generating a composite VLIW instruction, comprising:
inserting a scalar atom in the composite VLIW instruction to be executed at a single iteration;
inserting a vector atom in the composite VLIW instruction to be executed during all iterations indicated in a REPEAT field of the composite VLIW instruction; and
inserting branch and control atoms in the composite VLIW instruction to be executed at a single iteration.

### Claim at position 19 (depends on 19)

19: The method of claim 19, further comprising:
inserting in the scalar atom a field indicating an iteration in which the scalar atom is to be executed; and
inserting in the branch and control atoms a field indicating an iteration in which the branch and control atoms are to be executed.

### Claim at position 20 (depends on 13)

20: A programmable storage medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform the method of executing a very large instruction word (VLIW) instruction of claim 13.
