Energy efficient processor core architecture for image processor

Patent metadata

Field	Value
Publication number	US-9772852-B2
Application number	US-201514694815-A
Country	US
Kind code	B2
Filing date	Apr 23, 2015
Priority date	Apr 23, 2015
Publication date	Sep 26, 2017
Grant date	Sep 26, 2017

Abstract

An apparatus that includes a program controller to fetch and issue instructions is described. The apparatus includes an execution lane having at least one execution unit to execute the instructions. The execution lane is part of an execution lane array that is coupled to a two dimensional shift register array structure, wherein, execution lane s of the execution lane array are located at respective array locations and are coupled to dedicated registers at same respective array locations in the two-dimensional shift register array.

First claim

The invention claimed is:

1. An apparatus, comprising:
a program controller to fetch and issue instructions; and,
an execution lane array that is coupled to a two dimensional shift register array structure, wherein, execution lanes of the execution lane array are located at respective array locations and are coupled to dedicated registers at same respective array locations in the two-dimensional shift register array structure, wherein the execution lanes each comprise at least one execution unit to execute the instructions, the two dimensional shift register array structure comprising more rows and columns than the execution lane array.

2. The apparatus of claim 1 wherein the execution lanes are coupled to the program controller to receive and execute the instructions in SIMD fashion.

3. The apparatus of claim 2 wherein the execution lanes each comprise a memory unit.

4. The apparatus of claim 2 wherein the execution lanes each comprise a first ALU having an output coupled to a second ALU.

5. The apparatus of claim 2 wherein the instructions are contained in a larger data structure having a VLIW format.

6. The apparatus of claim 5 wherein the program controller is part of a scalar processor, the scalar processor to execute scalar instructions that are contained in the larger data structure having a VLIW format.

7. The apparatus of claim 6 wherein the scalar instructions include an instruction to broadcast a scalar value to the execution lanes of the execution lane array.

8. The apparatus of claim 6 wherein the scalar processor is coupled to its own dedicated random access memory.

9. The apparatus of claim 1 wherein different regions of the execution lane array are allocated with a different respective random access memory.

10. The apparatus of claim 9 wherein the different random access memories are to store register values that spill out of the two dimensional shift register array structure.

11. The apparatus of claim 9 wherein the different random access memories are to store look-up table information.

12. A non transitory machine readable storage medium having stored thereon a description of a circuit, the circuit comprising:
a program controller to fetch and issue instructions; and,
an execution lane array that is coupled to a two dimensional shift register array structure, wherein, execution lanes of the execution lane array are located at respective array locations and are coupled to dedicated registers at same respective array locations in the two-dimensional shift register array structure, wherein the execution lanes each comprise at least one execution unit to execute the instructions, the two dimensional shift register array structure comprising more rows and columns than the execution lane array.

13. The machine readable storage medium of claim 12 wherein the execution lanes are coupled to the program controller to receive and execute the instructions in SIMD fashion.

14. The machine readable storage medium of claim 13 wherein the execution lanes each comprise a memory unit.

15. The machine readable storage medium of claim 13 wherein the execution lanes each comprise a first ALU having an output coupled to a second ALU.

16. The machine readable storage medium of claim 13 wherein the instructions are contained in a larger data structure having a VLIW format.

17. The machine readable storage medium of claim 16 wherein the program controller is part of a scalar processor, the scalar processor to execute scalar instructions that are contained in the larger data structure having a VLIW format.

18. The machine readable storage medium of claim 12 wherein-different regions of the execution lane array are allocated with a different respective random access memory.

19. The machine readable storage medium of claim 18 wherein the different random access memories are to store register values that spill out of the two dimensional shift register array structure.

20. The machine readable storage medium of claim 18 wherein the different random access memories are to store look-up table information.

21. A computing system, comprising:
a plurality of processing cores coupled to a system memory;
an image processing unit, the image processing unit comprising:
a program controller to fetch and issue instructions; and,
an execution lane array that is coupled to a two dimensional shift register array structure, wherein, execution lanes of the execution lane array are located at respective array locations and are coupled to dedicated registers at same respective array locations in the two-dimensional shift register array structure, wherein the execution lanes each comprise at least one execution unit to execute the instructions, the two dimensional shift register array structure comprising more rows and columns than the execution lane array.

22. The computing system of claim 21 wherein the execution lanes of the execution lane array are coupled to the program controller to receive and execute the instructions in SIMD fashion.

23. The computing system of claim 21 wherein the execution lanes each comprise a memory unit.

24. The computing system of claim 21 wherein the execution lanes each comprise a first ALU having an output coupled to a second ALU.

25. The computing system of claim 21 wherein the instructions are contained in a larger data structure having a VLIW format.

26. The computing system of claim 21 wherein the program controller is part of a scalar processor, the scalar processor to execute scalar instructions that are contained in the larger data structure having a VLIW format.

27. The computing system of claim 21 wherein different regions of the execution lane are allocated with a different respective random access memory.

28. The computing system of claim 27 wherein the different random access memories are to store register values that spill out of the two dimensional shift register array structure.

Assignees

• Google Inc

Inventors

• Meixner Albert
• Redgrave Jason Rupert
• Shacham Ofer
• Finchelstein Daniel Frederic
• Zhu Qiuling

Classifications

• G06F9/3871

  Asynchronous instruction pipeline, e.g. using handshake signals between stages · CPC title

• G06F15/8023

  Two dimensional arrays, e.g. mesh, torus · CPC title

• G06F9/30134

  Register stacks; shift registers · CPC title

• H04N5/14

  Picture signal circuitry for video frequency region (cameras or camera modules comprising electronic image sensors, or control thereof H04N23/00) · CPC title

• G06F9/3802 Primary

  Instruction prefetching · CPC title