Techniques for data type conversion are described. An example uses an instruction that is to include fields for an opcode, an identification of source operand location, and an identification of destination operand location, wherein the opcode is to indicate instruction processing circuitry is to convert a 16-bit floating-point value from the identified source operand location into a 32-bit floating point value and store that 32-bit floating point value in one or more data element positions of the identified destination operand.

Techniques for data type conversion via instruction are described. An exemplary instruction is to include fields for an opcode, an identification of a source operand, and an identification of destination operand, wherein the opcode is to indicate instruction processing circuitry is to convert odd 16-bit floating point values from the identified source operand into 32-bit floating point values and store the 32-bit floating point values in data element positions of the identified destination operand.

One embodiment provides for a graphics processing unit (GPU) to accelerate machine learning operations, the GPU comprising an instruction cache to store a first instruction and a second instruction, the first instruction to cause the GPU to perform a floating-point operation, including a multi-dimensional floating-point operation, and the second instruction to cause the GPU to perform an integer operation; and a general-purpose graphics compute unit having a single instruction, multiple thread architecture, the general-purpose graphics compute unit to concurrently execute the first instruction and the second instruction.

An apparatus stores a program including a description of loop processing of iterating a plurality of instructions, and rearranges an execution sequence of the plurality of instructions in the program such that the loop processing is pipelined by software pipeline. The apparatus inserts an instruction to use a register for single instruction multiple data (SIMD) extension instruction, into the description of the loop processing in the program.

This invention is a streaming engine employed in a digital signal processor. A fixed data stream sequence is specified by a control register. The streaming engine fetches stream data ahead of use by a central processing unit and stores it in a stream buffer. Upon occurrence of a fault reading data from memory, the streaming engine identifies the data element triggering the fault preferably storing this address in a fault address register. The streaming engine defers signaling the fault to the central processing unit until this data element is used as an operand. If the data element is never used by the central processing unit, the streaming engine never signals the fault. The streaming engine preferably stores data identifying the fault in a fault source register. The fault address register and the fault source register are preferably extended control registers accessible only via a debugger.

In a method of operating a computer system, an instruction loop is executed by a processor in which each iteration of the instruction loop accesses a current data vector and an associated current vector predicate. The instruction loop is repeated when the current vector predicate indicates the current data vector contains at least one valid data element and the instruction loop is exited when the current vector predicate indicates the current data vector contains no valid data elements.

A method, apparatus and non-transitory computer readable medium are provided for permuting data registers to a target register. Two or more data registers are concatenated to form a concatenated data register. Each data register comprises a plurality of elements. A permutation instruction which uses one of the data registers as a data input register is executed and conditionally selects an element of the data input register by comparing a portion of an element of a pattern register to an immediate match field value. The selected element of the data input register is copied to an element in a target register at a position corresponding to a position of the element of the pattern register when the portion of the element of the pattern register matches the immediate match field value. When the portion of the element of the pattern register does not match, the target register remains unchanged.

A method is provided that includes receiving, in a permute network, a plurality of data elements for a vector instruction from a streaming engine, and mapping, by the permute network, the plurality of data elements to vector locations for execution of the vector instruction by a vector functional unit in a vector data path of a processor.

A Very Long Instruction Word (VLIW) digital signal processor particularly adapted for single instruction multiple data (SIMD) operation on various operand widths and data sizes. A vector compare instruction compares first and second operands and stores compare bits. A companion vector conditional instruction performs conditional operations based upon the state of a corresponding predicate data register bit. A predicate unit performs data processing operations on data in at least one predicate data register including unary operations and binary operations. The predicate unit may also transfer data between a general data register file and the predicate data register file.

Disclosed herein is a processor for deep learning. In one embodiment, the processor comprises: a load and store unit configured to load and store image pixel data and stencil data; a register unit, implementing a banked register file, configured to: load and store a subset of the image pixel data from the load and store unit, and concurrently provide access to image pixel values stored in a register file entry of the banked register file, wherein the subset of the image pixel data comprises the image pixel values stored in the register file entry; and a plurality of arithmetic logic units configured to concurrently perform one or more operations on the image pixel values stored in the register file entry and corresponding stencil data of the stencil data.