公开(公告)号：US20250117218A1

公开(公告)日：2025-04-10

申请号：US18927097

申请日：2024-10-25

Applicant: Intel Corporation

Inventor： Alexander Heinecke ,  Naveen Mellempudi ,  Robert Valentine ,  Mark Charney ,  Christopher Hughes ,  Evangelos Georganas ,  Zeev Sperber ,  Amit Gradstein ,  Simon Rubanovich

IPC: G06F9/30 ,  G06F9/38

Abstract: Techniques for converting FP16 to BF8 using bias are described. An exemplary embodiment utilizes 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 data from the identified first and second sources to packed bfloat8 data using bias terms from the identified source/destination operand and store the packed bfloat8 data 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 data from the identified first and second sources to packed bfloat8 data using bias terms from the identified source/destination operand and store the packed bfloat8 data into corresponding data element positions of the identified source/destination operand.

公开(公告)号：US20240248720A1

公开(公告)日：2024-07-25

申请号：US18627907

申请日：2024-04-05

Applicant: Intel Corporation

Inventor： Alexander Heinecke ,  Naveen Mellempudi ,  Robert Valentine ,  Mark Charney ,  Christopher Hughes ,  Evangelos Georganas ,  Zeev Sperber ,  Amit Gradstein ,  Simon Rubanovich

IPC: G06F9/30 ,  G06F7/499 ,  H03M7/24

CPC classification number: G06F9/30145 ,  G06F7/49947 ,  G06F9/30025 ,  G06F9/30036 ,  H03M7/24

Abstract: Techniques for converting FP16 data elements to BF8 data elements using a single instruction are described. An exemplary apparatus includes decoder circuitry to decode a single instruction, the single instruction to include a one or more fields to identify a source operand, one or more fields to identify a destination operand, and one or more fields for an opcode, the opcode to indicate that execution circuitry is to convert packed half-precision floating-point data from the identified source to packed bfloat8 data and store the packed bfloat8 data into corresponding data element positions of the identified destination operand; and execution circuitry to execute the decoded instruction according to the opcode to convert packed half-precision floating-point data from the identified source to packed bfloat8 data and store the packed bfloat8 data into corresponding data element positions.

公开(公告)号：US20240045691A1

公开(公告)日：2024-02-08

申请号：US17958374

申请日：2022-10-01

Applicant: Intel Corporation

Inventor： Naveen Mellempudi ,  Menachem Adelman ,  Evangelos Georganas ,  Amit Gradstein ,  Christopher Hughes ,  Alexander Heinecke ,  Simon Rubanovich ,  Uri Sherman ,  Zeev Sperber

IPC: G06F9/38 ,  G06F7/487 ,  G06F17/16 ,  G06F9/30

CPC classification number: G06F9/3802 ,  G06F7/4876 ,  G06F17/16 ,  G06F9/3001

Abstract: Systems, methods, and apparatuses relating to 8-bit floating-point matrix dot product instructions are described. A processor embodiment includes fetch circuitry to fetch an instruction having fields to specify an opcode and locations of a destination matrix having single-precision elements, a first source matrix, and a second source matrix, the source matrices having elements that each comprise a quadruple of 8-bit floating-point values, the opcode to indicate execution circuitry is to cause, for each element of the first source matrix and corresponding element of the second source matrix, a conversion of the 8-bit floating-point values to single-precision values, a multiplication of different pairs of converted single-precision values to generate plurality of results, and an accumulation of the results with previous contents of a corresponding element of the destination matrix, decode circuitry to decode the fetched instruction, and the execution circuitry to respond to the decoded instruction as specified by the opcode.

公开(公告)号：US20240045687A1

公开(公告)日：2024-02-08

申请号：US17958365

申请日：2022-10-01

Applicant: Intel Corporation

Inventor： Alexander Heinecke ,  Menachem Adelman ,  Evangelos Georganas ,  Amit Gradstein ,  Christopher Hughes ,  Naveen Mellempudi ,  Simon Rubanovich ,  Uri Sherman ,  Zeev Sperber

IPC: G06F9/30

CPC classification number: G06F9/3016 ,  G06F9/3013

Abstract: Techniques for FP8 classification or manipulation using single instructions are described. An exemplary instruction includes fields for an opcode, an identification of a location of a packed data source operand, an indication of one or more classification checks to perform, and an identification of a packed data destination operand, wherein the opcode is to indicate that execution circuitry is to perform, for each data element position of the packed data source operand, a classification according to the indicated one or more classification checks and store a result of the classification in a corresponding data element position of the destination operand.

公开(公告)号：US20240045682A1

公开(公告)日：2024-02-08

申请号：US17958370

申请日：2022-10-01

Applicant: Intel Corporation

Inventor： Alexander Heinecke ,  Menachem Adelman ,  Evangelos Georganas ,  Amit Gradstein ,  Christopher Hughes ,  Naveen Mellempudi ,  Simon Rubanovich ,  Uri Sherman ,  Zeev Sperber

IPC: G06F9/30

CPC classification number: G06F9/30145 ,  G06F9/30036 ,  G06F9/3001

Abstract: Techniques for scale and reduction of FP8 data elements are described. An exemplary instruction includes fields for an having fields for an opcode, an identification of a location of a first packed data source operand, an identification of a location of a second packed data source operand, and an identification of a packed data destination operand, wherein the opcode is to indicate that execution circuitry is to perform, for each data element position of the packed data source operands, a floating point scale operation of a FP8 data element of the first packed data source by multiplying the data element by a power of 2 value, wherein a value of the exponent of the power of 2 value is a floor value of a FP8 data element of the second packed data source, and store a result of the floating point scale operation into a corresponding data element position of the packed data destination operand.

公开(公告)号：US20230315450A1

公开(公告)日：2023-10-05

申请号：US18313026

申请日：2023-05-05

Applicant: Intel Corporation

Inventor： Naveen Mellempudi ,  Alexander F. Heinecke ,  Robert Valentine ,  Mark J. Charney ,  Christopher J. Hughes ,  Evangelos Georganas ,  Zeev Sperber ,  Amit Gradstein ,  Simon Rubanovich

IPC: G06F9/30 ,  G06F7/499 ,  G06F9/38

CPC classification number: G06F9/30036 ,  G06F7/49915 ,  G06F9/30196 ,  G06F9/3887

Abstract: Systems, methods, and apparatuses relating to 8-bit floating-point matrix dot product instructions are described. A processor embodiment includes fetch circuitry to fetch an instruction having fields to specify an opcode and locations of a destination matrix having single-precision elements, a first source matrix, and a second source matrix, the source matrices having elements that each comprise a quadruple of 8-bit floating-point values, the opcode to indicate execution circuitry is to cause, for each element of the first source matrix and corresponding element of the second source matrix, a conversion of the 8-bit floating-point values to single-precision values, a multiplication of different pairs of converted single-precision values to generate plurality of results, and an accumulation of the results with previous contents of a corresponding element of the destination matrix, decode circuitry to decode the fetched instruction, and the execution circuitry to respond to the decoded instruction as specified by the opcode.

公开(公告)号：US11501139B2

公开(公告)日：2022-11-15

申请号：US15869582

申请日：2018-01-12

Applicant: Intel Corporation

Inventor： Naveen Mellempudi ,  Dipankar Das

IPC: G06T1/20 ,  G06F5/01 ,  G06N3/063 ,  G06F7/487 ,  G06F7/544 ,  G06N3/04 ,  G06N3/08

Abstract: One embodiment provides for a machine-learning accelerator device a multiprocessor to execute parallel threads of an instruction stream, the multiprocessor including a compute unit, the compute unit including a set of functional units, each functional unit to execute at least one of the parallel threads of the instruction stream. The compute unit includes compute logic configured to execute a single instruction to scale an input tensor associated with a layer of a neural network according to a scale factor, the input tensor stored in a floating-point data type, the compute logic to scale the input tensor to enable a data distribution of data of the input tensor to be represented by a 16-bit floating point data type.

公开(公告)号：US20220206801A1

公开(公告)日：2022-06-30

申请号：US17134373

申请日：2020-12-26

Applicant: Intel Corporation

Inventor： Naveen Mellempudi ,  Alexander F. Heinecke ,  Robert Valentine ,  Mark J. Charney ,  Christopher J. Hughes ,  Evangelos Georganas ,  Zeev Sperber ,  Amit Gradstein ,  Simon Rubanovich

IPC: G06F9/30 ,  G06F7/499 ,  G06F9/38

Abstract: Systems, methods, and apparatuses relating to 8-bit floating-point matrix dot product instructions are described. A processor embodiment includes fetch circuitry to fetch an instruction having fields to specify an opcode and locations of a destination matrix having single-precision elements, a first source matrix, and a second source matrix, the source matrices having elements that each comprise a quadruple of 8-bit floating-point values, the opcode to indicate execution circuitry is to cause, for each element of the first source matrix and corresponding element of the second source matrix, a conversion of the 8-bit floating-point values to single-precision values, a multiplication of different pairs of converted single-precision values to generate plurality of results, and an accumulation of the results with previous contents of a corresponding element of the destination matrix, decode circuitry to decode the fetched instruction, and the execution circuitry to respond to the decoded instruction as specified by the opcode.

公开(公告)号：US11321805B2

公开(公告)日：2022-05-03

申请号：US17083588

申请日：2020-10-29

Applicant: Intel Corporation

Inventor： Naveen Mellempudi ,  Dheevatsa Mudigere ,  Dipankar Das ,  Srinivas Sridharan

IPC: G06T1/20 ,  G06N3/063 ,  G06F17/16 ,  G06F7/523 ,  G06F5/01 ,  G06F7/501 ,  G06F17/15 ,  G06N3/04 ,  G06F7/544 ,  G06N3/08

Abstract: One embodiment provides for a graphics processing unit to perform computations associated with a neural network, the graphics processing unit comprising compute unit including a hardware logic unit having dynamic precision fixed-point logic, the compute unit to receive a set of dynamic fixed-point tensors, compute, via the dynamic precision fixed-point logic, a right-shift value using an absolute maximum value within the set of dynamic fixed-point tensors and a dynamic range of the set of dynamic fixed-point tensors, right-shift data values within the set of dynamic fixed-point tensors based on the right-shift value, increment a shared exponent associated with the set of dynamic fixed-point tensors based on the right-shift value, perform a compute operation on the set of dynamic fixed-point tensors, and generate an output tensor via the compute operation on the set of dynamic fixed-point tensors.

公开(公告)号：US10719317B2

公开(公告)日：2020-07-21

申请号：US16003555

申请日：2018-06-08

Applicant: INTEL CORPORATION

Inventor： Victor Lee ,  Ugonna Echeruo ,  George Chrysos ,  Naveen Mellempudi

IPC: G06F9/30

Abstract: Methods and apparatuses relating to a vector instruction with a register operand with an elemental offset are described. In one embodiment, a hardware processor includes a decode unit to decode a vector instruction with a register operand with an elemental offset to access a first number of elements in a register specified by the register operand, wherein the first number is a total number of elements in the register minus the elemental offset, access a second number of elements in a next logical register, wherein the second number is the elemental offset, and combine the first number of elements and the second number of elements as a data vector, and an execution unit to execute the vector instruction on the data vector.