Instructions to convert from FP16 to BF8

1. An apparatus comprising:

decoder circuitry to decode a single instruction, the single instruction to include one or more fields to identify a first source operand, one or more fields to identify a second source operand, one or more fields to identify a source/destination operand, and one or more fields for an opcode, wherein the opcode is to indicate that execution circuitry is to convert packed half-precision floating-point data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand; and

execution circuitry to execute the decoded instruction according to the opcode to convert packed half-precision floating-point data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand.

2. The apparatus of claim 1 , wherein the first and second source operands are vector registers.

3. The apparatus of claim 1 , wherein the bias terms are 8-bit values.

4. The apparatus of claim 1 , wherein when a half-precision floating-point data value is infinite, a bfloat8 data value is set to be bits 15:8 of the half-precision floating-point data value.

5. The apparatus of claim 1 , wherein when a half-precision floating-point data value is not-a-number, a bfloat8 data value has bits 7:2 set to be bits 15:10 of the half-precision floating-point data value, bit 1 set to be a value of 1, and bit 0 set to be bit 8 of the half-precision floating-point data value.

6. The apparatus of claim 1 , wherein when a half-precision floating-point data value is not-a-number, a bfloat8 data value is to be computed by setting a first temporary value to be 0, setting bits 7:0 of the first temporary value to be a bias term, setting a second temporary value to be the half-precision floating-point data value plus the first temporary value, and the bfloat8 data value is set to be bits 15:8 of the second temporary value.

7. The apparatus of claim 1 , wherein the single instruction is further to include one or more fields to identify a writemask operand, wherein one or more bits of the writemask operand are to indicate to execution circuitry which of the converted bfloat8 data values are to be written in the source/destination operand.

8. A method comprising:

decoding a single instruction, the single instruction to include one or more fields to identify a first source operand, one or more fields to identify a second source operand, one or more fields to identify a source/destination operand, and one or more fields for an opcode, wherein the opcode is to indicate that execution circuitry is to convert packed half-precision floating-point data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand; and

executing the decoded instruction according to the opcode to convert packed half-precision floating-point data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand.

9. The method of claim 8 , wherein the one or more fields to identify the first source operand is to identify a vector register.

10. The method of claim 8 , wherein the bias terms are 8-bit values.

11. The method of claim 8 , wherein when a half-precision floating-point data value is infinite, a bfloat8 data value is set to be bits 15:8 of the half-precision floating-point data value.

12. The method of claim 8 , wherein when a half-precision floating-point data value is mot-a-number, a bfloat8 data value has bits 7:2 set to be bits 15:10 of the half-precision floating-point data value, bit 1 set to be a value of 1, and bit 0 set to be bit 8 of the half-precision floating-point data value.

13. The method of claim 8 , wherein when a half-precision floating-point data value is not-a-number, a bfloat8 data value is to be computed by setting a first temporary value to be 0, setting bits 7:0 of the first temporary value to be a bias term, setting a second temporary value to be the half-precision floating-point data value plus the first temporary value, and the bfloat8 data is set to be bits 15:8 of the second temporary value.

14. The method of claim 8 , wherein the single instruction is further to include one or more fields to identify a writemask operand, wherein one or more bits of the writemask operand are to indicate to execution circuitry which of the converted bfloat8 data values are to be written in the source/destination operand.

15. The method of claim 8 , further comprising translating the single instruction into one or more instructions of a different instruction set architecture prior to decoding, wherein executing of the one or more instructions of the different instruction set architecture is to be functionally equivalent as the executing according to the opcode of the single instruction.

16. A non-transitory machine-readable medium storing an instance of a single instruction that includes one or more fields to identify a first source operand, one or more fields to identify a second source operand, one or more fields to identify a source/destination operand, and one or more fields for an opcode, wherein the opcode is to indicate that execution circuitry is to convert packed half-precision data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand, wherein the instance of the single instruction is to be handled by a processor by performing a method, the method comprising:

decoding the single instruction; and

executing the decoded instruction according to the opcode to convert packed half-precision data values from the identified first and second source operands to packed bfloat8 data values using bias terms from the identified source/destination operand and store the packed bfloat8 data values into corresponding data element positions of the identified source/destination operand.

17. The non-transitory machine-readable medium of claim 16 , wherein when a half-precision floating-point data value is infinite, a bfloat8 data value is set to be bits 15:8 of the half-precision floating-point data value.

18. The non-transitory machine-readable medium of claim 16 , wherein when a half-precision floating-point data value is not-a-number, a bfloat8 data value has bits 7:2 set to be bits 15:10 of the half-precision floating-point data value, bit 1 set to be a value of 1, and bit 0 set to be bit 8 of the half-precision floating-point data value.

19. The non-transitory machine-readable medium of claim 16 , wherein when a half-precision floating-point data value is not-a-number, a bfloat8 data value is to be computed by setting a first temporary value to be 0, setting bits 7:0 of the first temporary value to be a bias term, setting a second temporary value to be the half-precision floating-point data value plus the first temporary value, and the bfloat8 data value is set to be bits 15:8 of the second temporary value.

20. The non-transitory machine-readable medium of claim 16 , wherein the field for the identifier of the first source operand is to identify a vector register.