Systems, apparatuses, and methods for performing a butterfly horizontal and cross add or substract in response to a single instruction

Patent metadata

Field	Value
Publication number	US-9459865-B2
Application number	US-201113992236-A
Country	US
Kind code	B2
Filing date	Dec 23, 2011
Priority date	Dec 23, 2011
Publication date	Oct 4, 2016
Grant date	Oct 4, 2016

Abstract

Embodiments of systems, apparatuses, and methods for performing in a computer processor vector packed butterfly horizontal cross add or subtract of packed data elements in response to a single vector packed butterfly horizontal cross add or subtract instruction that includes a destination vector register operand, a source vector register operand, an immediate, and an opcode are described.

First claim

What is claimed is:

1. A method comprising:
executing a single instruction to, for each data lane of a source vector register, calculate a horizontal and cross addition or subtraction between packed data elements of the source vector register where the decision of whether to add or subtract is dependent upon a bit in an immediate value of the instruction, wherein the instruction includes a destination vector register operand, the source vector register operand, the immediate, and an opcode; and
storing a result of each a horizontal and cross addition or subtraction between packed data elements into the destination register, wherein for the least significant data element position of the destination register, the result stored is the least significant data element of the source register being added to or subtracted from the second most least significant data element of the source register wherein the decision of addition or subtraction is based on the least significant bit of the immediate;
wherein for the second most least significant data element position of the destination register, the result stored is the result of the third most least significant data element of the source register being added to or subtracted from the fourth most least significant data element of the source register wherein the decision of addition or subtraction is based on the third most least significant bit of the immediate;
wherein for the third most least significant data element position of the destination register, the result stored is the result of the second most least significant data element of the source register being added to or subtracted from the least significant data element of the source register wherein the decision of addition or subtraction is based on the second most least significant bit of the immediate; and
wherein for the fourth most significant data element position of the destination register, the result stored is the result of the fourth most least significant data element of the source register being added to or subtracted from the third most least significant data element of the source register wherein the decision of addition or subtraction is based on the fourth most least significant bit of the immediate.

2. The method of claim 1 , wherein there are a plurality of data lanes.

3. The method of claim 1 , wherein a number of data lanes to be processed is dependent upon the size of the destination vector register.

4. The method of claim 1 , wherein the source and destination vector registers are 128-bit, 256-bit, or 512-bit in size.

5. The method of claim 1 , wherein the packed data elements of the source and destination registers are 8-bit, 16-bit, 32-bit, or 64-bit in size.

6. The method of claim 5 , wherein the size of the packed data elements of the source is defined by the opcode.

7. The method of claim 1 , wherein the immediate is an 8-bit value.

8. An article of manufacture comprising:
a non-transitory machine-readable storage medium having stored thereon an occurrence of a single instruction, wherein the instruction's format specifies as its source operands a vector register and an immediate and specifies as its destination a single destination vector register, and wherein the instruction format includes an opcode which instructs a machine, responsive to the single occurrence of the single instruction, to cause for each data lane of the source vector register, a calculation of a horizontal and cross addition or subtraction between packed data elements of the source vector register where the decision of whether to add or subtract is dependent upon a bit in the immediate value of the instruction, and storage of a result of each a horizontal and cross addition or subtraction between packed data elements results into the destination register,
wherein for the least significant data element position of the destination register, the result stored is the least significant data element of the source register being added to or subtracted from the second most least significant data element of the source register wherein the decision of addition or subtraction is based on the least significant bit of the immediate;
wherein for the second most least significant data element position of the destination register, the result stored is the result of the third most least significant data element of the source register being added to or subtracted from the fourth most least significant data element of the source register wherein the decision of addition or subtraction is based on the third most least significant bit of the immediate;
wherein for the third most least significant data element position of the destination register, the result stored is the result of the second most least significant data element of the source register being added to or subtracted from the least significant data element of the source register wherein the decision of addition or subtraction is based on the second most least significant bit of the immediate; and
wherein for the fourth most significant data element position of the destination register, the result stored is the result of the fourth most least significant data element of the source register being added to or subtracted from the third most least significant data element of the source register wherein the decision of addition or subtraction is based on the fourth most least significant bit of the immediate.

9. The article of manufacture of claim 8 , wherein there are a plurality of data lanes.

10. The article of manufacture of claim 8 , wherein a number of data lanes to be processed is dependent upon the size of the destination vector register.

11. The article of manufacture of claim 8 , wherein the source and destination vector registers are 128-bit, 256-bit, or 512-bit in size.

12. The article of manufacture of claim 8 , wherein the packed data elements of the source and destination registers are 8-bit, 16-bit, 32-bit, or 64-bit in size.

13. The article of manufacture of claim 8 , wherein the size of the packed data elements of the source is defined by the opcode.

14. The article of manufacture of claim 8 , wherein the immediate is an 8-bit value.

15. An apparatus comprising;
a hardware decoder to decode a single instruction that includes a destination vector register operand, a source vector register operand, an immediate, and an opcode;
execution circuitry to, for each data lane of the source vector register, calculate a horizontal and cross addition or subtraction between packed data elements of the source vector register where the decision of whether to add or subtract is dependent upon a bit in the immediate value of the instruction, and store a result for each a horizontal and cross addition or subtraction between packed data elements results into the destination register, wherein for the least significant data element position of the destination register, the result stored is the least significant data element of the source register being added to or subtracted from the second most least significant data element of the source register wherein the decision of addition or subtraction is based on the least significant bit of the immediate;
wherein for the second most least significant data element position of the destination register, the result stored is the result of the third most least significant data element of the source register being added to or subtracted from the fourth most least significant data element of the source register wherein the decision of addition or subtraction is based on the third most least significant bit of the immediate;
wherein for the third most least

Assignees

• Ould-Ahmed-Vall Elmoustapha
• Hagog Mostafa
• Valentine Robert
• Gradstein Amit
• Rubanovich Simon
• Sperber Zeev
• Intel Corp

Inventors

• Ould-Ahmed-Vall Elmoustapha
• Hagog Mostafa
• Valentine Robert
• Gradstein Amit
• Rubanovich Simon
• Sperber Zeev

Classifications

• G06F9/30145

  Instruction analysis, e.g. decoding, instruction word fields · CPC title

• G06F9/30018

  Bit or string instructions · CPC title

• G06F9/30014

  with variable precision · CPC title

• G06F17/142

  Fast Fourier transforms, e.g. using a Cooley-Tukey type algorithm · CPC title

• G06F9/30036

  Instructions to perform operations on packed data, e.g. vector, tile or matrix operations · CPC title