Abstract: A method of an aspect includes receiving a floating point rounding instruction. The floating point rounding instruction indicates a source of one or more floating point data elements, indicates a number of fraction bits after a radix point that each of the one or more floating point data elements are to be rounded to, and indicates a destination storage location. A result is stored in the destination storage location in response to the floating point rounding instruction. The result includes one or more rounded result floating point data elements. Each of the one or more rounded result floating point data elements includes one of the floating point data elements of the source, in a corresponding position, which has been rounded to the indicated number of fraction bits. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: Instructions and logic provide SIMD vector packed tuple cross-comparison functionality. Some processor embodiments include first and second registers with a variable plurality of data fields, each of the data fields to store an element of a first data type. The processor executes a SIMD instruction for vector packed tuple cross-comparison in some embodiments, which for each data field of a portion of data fields in a tuple of the first register, compares its corresponding element with every element of a corresponding portion of data fields in a tuple of the second register and sets a mask bit corresponding to each element of the second register portion, in a bit-mask corresponding to each unmasked element of the corresponding first register portion, according to the corresponding comparison. In some embodiments bit-masks are shifted by corresponding elements in data fields of a third register. The comparison type is indicated by an immediate operand.

Abstract: Disclosed embodiments relate to computing dot products of nibbles in tile operands. In one example, a processor includes decode circuitry to decode a tile dot product instruction having fields for an opcode, a destination identifier to identify a M by N destination matrix, a first source identifier to identify a M by K first source matrix, and a second source identifier to identify a K by N second source matrix, each of the matrices containing doubleword elements, and execution circuitry to execute the decoded instruction to perform a flow K times for each element (M,N) of the identified destination matrix to generate eight products by multiplying each nibble of a doubleword element (M,K) of the identified first source matrix by a corresponding nibble of a doubleword element (K,N) of the identified second source matrix, and to accumulate and saturate the eight products with previous contents of the doubleword element (M,N).

Abstract: Disclosed embodiments relate to systems and methods for performing instructions to transpose rectangular tiles. In one example, a processor includes fetch circuitry to fetch an instruction having fields to specify an opcode and locations of first destination, second destination, first source, and second source matrices, the specified opcode to cause the processor to process each of the specified source and destination matrices as a rectangular matrix, decode circuitry to decode the fetched rectangular matrix transpose instruction, and execution circuitry to respond to the decoded rectangular matrix transpose instruction by transposing each row of elements of the specified first source matrix into a corresponding column of the specified first destination matrix and transposing each row of elements of the specified second source matrix into a corresponding column of the specified second destination matrix.

Abstract: Embodiments detailed herein relate to systems and methods to store a tile register pair to memory. In one example, a processor includes: decode circuitry to decode a store matrix pair instruction having fields for an opcode and source and destination identifiers to identify source and destination matrices, respectively, each matrix having a PAIR parameter equal to TRUE; and execution circuitry to execute the decoded store matrix pair instruction to store every element of left and right tiles of the identified source matrix to corresponding element positions of left and right tiles of the identified destination matrix, respectively, wherein the executing stores a chunk of C elements of one row of the identified source matrix at a time.

Abstract: Embodiments detailed herein relate to systems and methods to zero a tile register pair. In one example, a processor includes decode circuitry to decode a matrix pair zeroing instruction having fields for an opcode and an identifier to identify a destination matrix having a PAIR parameter equal to TRUE; and execution circuitry to execute the decoded matrix pair zeroing instruction to zero every element of a left matrix and a right matrix of the identified destination matrix.

Abstract: Embodiments detailed herein relate to matrix operations. For example, embodiments of instruction support for matrix (tile) dot product operations are detailed. Exemplary instructions including computing a dot product of signed words and accumulating in a quadword data elements of a matrix pair. Additionally, in some instances, non-accumulating quadword data elements of the matrix pair are set to zero.

Abstract: Embodiments detailed herein relate to matrix (tile) operations. For example, decode circuitry to decode an instruction having fields for an opcode and a memory address, and execution circuitry to execute the decoded instruction to store configuration information about usage of storage for two-dimensional data structures at the memory address.

Abstract: Embodiments detailed herein relate to systems and methods to load a tile register pair. In one example, a processor includes: decode circuitry to decode a load matrix pair instruction having fields for an opcode and source and destination identifiers to identify source and destination matrices, respectively, each matrix having a PAIR parameter equal to TRUE; and execution circuitry to execute the decoded load matrix pair instruction to load every element of left and right tiles of the identified destination matrix from corresponding element positions of left and right tiles of the identified source matrix, respectively, wherein the executing operates on one row of the identified destination matrix at a time, starting with the first row.

Abstract: A processor for floating point underflow detection includes circuitry to decode a first instruction and a floating point unit. The decoded instruction, when executed by the processor, may be for performing a fused multiply-add (FMA) operation. The floating point unit includes circuitry to determine a non-normalized result of the first instruction based on a first input, a second input, and a third input. The floating point unit further includes circuitry to determine whether underflow exists in the non-normalized result based on a first exponent of the first input, a second exponent of the second input, and a third exponent of the third input.

Abstract: An apparatus and method are described for performing vector compression. For example, one embodiment of a processor comprises: vector compression logic to compress a source vector comprising a plurality of valid data elements and invalid data elements to generate a destination vector in which valid data elements are stored contiguously on one side of the destination vector, the vector compression logic to utilize a bit mask associated with the source vector and comprising a plurality of bits, each bit corresponding to one of the plurality of data elements of the source vector and indicating whether the data element comprises a valid data element or an invalid data element, the vector compression logic to utilize indices of the bit mask and associated bit values of the bit mask to generate a control vector; and shuffle logic to shuffle/permute the data elements of the source vector to the destination vector in accordance with the control vector.

Abstract: An apparatus is described having an instruction execution pipeline that has a vector functional unit to support a vector multiply add instruction. The vector multiply add instruction to multiply respective K bit elements of two vectors and accumulate a portion of each of their respective products with another respective input operand in an X bit accumulator, where X is greater than K.

Abstract: A method of an aspect includes receiving a floating point scaling instruction. The floating point scaling instruction indicates a first source including one or more floating point data elements, a second source including one or more corresponding floating point data elements, and a destination. A result is stored in the destination in response to the floating point scaling instruction. The result includes one or more corresponding result floating point data elements each including a corresponding floating point data element of the second source multiplied by a base of the one or more floating point data elements of the first source raised to a power of an integer representative of the corresponding floating point data element of the first source. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: A method of an aspect includes receiving a floating point round-off amount determination instruction. The instruction indicates a source of one or more floating point data elements, indicates a number of fraction bits after a radix point, and indicates a destination storage location. A result including one or more result floating point data elements is stored in the destination storage location in response to the floating point round-off amount determination instruction. Each of the one or more result floating point data elements includes a difference between a corresponding floating point data element of the source in a corresponding position, and a rounded version of the corresponding floating point data element of the source that has been rounded to the indicated number of the fraction bits. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: A method of an aspect includes receiving a floating point scaling instruction. The floating point scaling instruction indicates a first source including one or more floating point data elements, a second source including one or more corresponding floating point data elements, and a destination. A result is stored in the destination in response to the floating point scaling instruction. The result includes one or more corresponding result floating point data elements each including a corresponding floating point data element of the second source multiplied by a base of the one or more floating point data elements of the first source raised to a power of an integer representative of the corresponding floating point data element of the first source. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: A method of an aspect includes receiving a floating point scaling instruction. The floating point scaling instruction indicates a first source including one or more floating point data elements, a second source including one or more corresponding floating point data elements, and a destination. A result is stored in the destination in response to the floating point scaling instruction. The result includes one or more corresponding result floating point data elements each including a corresponding floating point data element of the second source multiplied by a base of the one or more floating point data elements of the first source raised to a power of an integer representative of the corresponding floating point data element of the first source. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: A method of an aspect includes receiving a floating point scaling instruction. The floating point scaling instruction indicates a first source including one or more floating point data elements, a second source including one or more corresponding floating point data elements, and a destination. A result is stored in the destination in response to the floating point scaling instruction. The result includes one or more corresponding result floating point data elements each including a corresponding floating point data element of the second source multiplied by a base of the one or more floating point data elements of the first source raised to a power of an integer representative of the corresponding floating point data element of the first source. Other methods, apparatus, systems, and instructions are disclosed.

Abstract: An example processor includes a register and a fused multiply-add (FMA) low functional unit. The register stores first, second, and third floating point (FP) values. The FMA low functional unit receives a request to perform an FMA low operation: multiplies the first FP value with the second FP value to obtain a first product value; adds the first product with the third FP value to generate a first result value; rounds the first result to generate a first FMA value; multiplies the first FP value with the second FP value to obtain a second product value; adds the second product value with the third FP value to generate a second result value; and subtracts the FMA value from the second result value to obtain a third result value, which can then be normalized and rounded (FMA low result) and sent the FMA low result to an application.

Abstract: An example processor includes a register and an ADD low functional unit. The register stores first, second, and third floating point (FP) values. The ADD low functional unit receives a request to perform an ADD low operation and, responsive to the request: adds the first FP value with the second FP value to obtain a first sum value; rounds the first sum value to generate an ADD value; adds the first FP value with the second FP value to obtain a second sum value; subtracts the ADD value from the second sum value to generate a difference value; normalizes the difference value to obtain a normalized difference value; rounds the normalized difference value to generate an ADD low value; and sends the ADD low value to an application.

Abstract: A processor for floating point underflow detection includes circuitry to decode a first instruction and a floating point unit. The decoded instruction, when executed by the processor, may be for performing a fused multiply-add (FMA) operation. The floating point unit includes circuitry to determine a non-normalized result of the first instruction based on a first input, a second input, and a third input. The floating point unit further includes circuitry to determine whether underflow exists in the non-normalized result based on a first exponent of the first input, a second exponent of the second input, and a third exponent of the third input.