Instruction execution that broadcasts and masks data values at different levels of granularity

• US 9,424,327 B2
• Filed: 12/23/2011
• Issued: 08/23/2016
• Est. Priority Date: 12/23/2011
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Ould-Ahmed-Vall, Elmoustapha, Valentine, Robert, Corbal, Jesus, Toll, Bret L., Charney, Mark J.
• Primary Examiner(s): Beausoliel, Jr., Robert
• Assistant Examiner(s): Pham, Michael
• Application Number:
  US13/976,433
• Publication Number:
  US 20130290254A1
• Time in Patent Office: 1,705 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: G06F 21/6227 where protection concerns the structure of data, e.g. records, types, queries
• CPC Class Codes: G06F 21/6254 by anonymising data, e.g. decorrelating personal data from the owner's identification
• CPC Class Codes: G06F 21/70 Protecting specific internal or peripheral components, in which the protection of a component leads to protection of the entire computer
• CPC Class Codes: G06F 9/30018 Bit or string instructions
• CPC Class Codes: G06F 9/30032 Movement instructions, e.g. MOVE, SHIFT, ROTATE, SHUFFLE
• CPC Class Codes: G06F 9/30036 Instructions to perform operations on packed data, e.g. vector, tile or matrix operations
• CPC Class Codes: G06F 9/30101 Special purpose registers
• CPC Class Codes: G06F 9/3802 Instruction prefetching
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

An apparatus is described that includes an execution unit to execute a first instruction and a second instruction. The execution unit includes input register space to store a first data structure to be replicated when executing the first instruction and to store a second data structure to be replicated when executing the second instruction. The first and second data structures are both packed data structures. Data values of the first packed data structure are twice as large as data values of the second packed data structure. The execution unit also includes replication logic circuitry to replicate the first data structure when executing the first instruction to create a first replication data structure, and, to replicate the second data structure when executing the second data instruction to create a second replication data structure. The execution unit also includes masking logic circuitry to mask the first replication data structure at a first granularity and mask the second replication data structure at a second granularity. The second granularity is twice as fine as the first granularity.

Claims

1. An apparatus, comprising:

an execution unit to execute a first instruction and a second instruction, said execution unit including the following;

input register space to store a first data structure to be replicated when executing said first instruction and to store a second data structure to be replicated when executing said second instruction, said first and second data structures both being packed data structures, data values of said first packed data structure being twice as large as data values of said second packed data structure;

replication logic circuitry to replicate said first data structure when executing said first instruction to create a first replication data structure, and, to replicate said second data structure when executing said second instruction to create a second replication data structure; and

masking logic circuitry to mask said first replication data structure at a first granularity and mask said second replication data structure at a second granularity, said second granularity being twice as fine as said first granularity.

2. The apparatus of claim 1 wherein said first replication data structure comprises four copies of said first data structure.

3. The apparatus of claim 1 wherein said first replication data structure comprises two copies of said first data structure.

4. The apparatus of claim 1 wherein said second replication data structure comprises eight copies of said second data structure.

5. The apparatus of claim 1 wherein said second replication data structure comprises four copies of said second data structure.

6. The apparatus of claim 1 wherein said second replication data structure comprises two copies of said second data structure.

7. The apparatus of claim 1 wherein said data values of said first packed data structure are 64 bits each and said data values of said second packed data structure are 32 bits each.

8. A method, comprising:

fetching a first data structure for a first instruction, said first data structure being a first packed data structure;

executing said first instruction by replicating said first data structure within an execution unit to create a first replication data structure and masking said first replication data structure;

fetching a second data structure for a second instruction, said second data structure being a second packed data structure, data values of said first packed data structure being twice as large as data values of said second packed data structure; and

executing said second instruction by replicating said second data structure within said execution unit to create a second replication data structure and masking said second replication data structure at a granularity that is twice as fine as said masking of said first replication data structure.

9. The method of claim 8 wherein said first and second data values are floating point data values.

10. The method of claim 9 wherein said first data values are 64 bits and said second data values are 32 bits.

11. The method of claim 8 wherein said first replication data structure contains any of the following number of copies of said first data structure:

two; and

four.

12. The method of claim 11 wherein said second replication data structure contains any of the following number of copies of said second data structure:

two;

four; and

eight.

13. The method of claim 8 wherein said first packed data structure has any of the following number of data values:

two; and

four.

14. The method of claim 13 wherein said second packed data structure has any of the following number of data values:

two;

four; and

eight.

15. An apparatus, comprising:

an execution unit to execute a first instruction and a second instruction, said execution unit including the following;

input register space to store a respective data structure to be replicated for each of a plurality of instructions to be executed by said execution unit, said respective data structure for each of said plurality of instructions including;

a first unpacked data structure for a first instruction;

a second unpacked data structure for a second instruction where the second unpacked data structure is twice as large as said first unpacked data structure, a third packed data structure for a third instruction whose data values are the same size as said first unpacked data structure, a fourth packed data structure for a fourth instruction whose data values are the same size as said second unpacked data structure;

replication logic circuitry to;

replicate said first unpacked data structure when executing said first instruction to create a first replication data structure, replicate said second unpacked data structure when executing said second instruction to create a second replication data structure, replicate said third packed data structure when executing said third instruction to create a third replication data structure, replicate said fourth packed data structure when executing said fourth instruction to create a fourth replication data structure; and

masking logic circuitry to mask said first replication data structure and said third replication data structure at a first granularity and mask said second replication data structure and said fourth replication data structure at a second granularity, said first granularity being twice as fine as said second granularity.

16. The apparatus of claim 15 wherein said replication of said fourth instruction produces the fourth replication data structure containing any of the following number of copies of said fourth packed data structure:

two; and

four.

17. The apparatus of claim 16 wherein said replication of said third instruction produces the third replication data structure containing any of the following number of copies of said third packed data structure:

two;

four; and

eight.

18. The apparatus of claim 15 wherein said fourth packed data structure has any of the following number of data values:

two; and

four.

19. The apparatus of claim 18 wherein said third packed data structure has any of the following number of data values:

two;

four; and

eight.

20. The apparatus of claim 15 wherein said first unpacked data structure is 32 bits and said second unpacked data structure is 64 bits.