Aspects for matrix multiplication in neural network are described herein. The aspects may include a master computation module configured to receive a first matrix and transmit a row vector of the first matrix. In addition, the aspects may include one or more slave computation modules respectively configured to store a column vector of a second matrix, receive the row vector of the first matrix, and multiply the row vector of the first matrix with the stored column vector of the second matrix to generate a result element. Further, the aspects may include an interconnection unit configured to combine the one or more result elements generated respectively by the one or more slave computation modules to generate a row vector of a result matrix and transmit the row vector of the result matrix to the master computation module.

Systems and methods selectively bypass address-generation hardware in processor instruction pipelines. In an embodiment, a processor includes an address-generation stage and an address-generation-bypass-determination unit (ABDU). The ABDU receives a load/store instruction. If an effective address for the load/store instruction is not known at the ABDU, the ABDU routes the load/store instruction via the address-generation stage of the processor. If, however, the effective address of the load/store instruction is known at the ABDU, the ABDU routes the load/store instruction to bypass the address-generation stage of the processor.

Aspects for matrix multiplication in neural network are described herein. The aspects may include a controller unit configured to receive a matrix-multiply-matrix (MM) instruction that includes a first starting address of a first matrix, a first size of the first matrix, a second starting address of a second matrix, and a second size of the second matrix; multiple computation modules configured to respectively multiply, in response to the MM instruction, row vectors of the first matrix with column vectors of the second matrix to generate one or more result elements; and an interconnection unit configured to combine the result elements to generate one or more row vectors of a result matrix.

Predicting a Table of Contents (TOC) pointer value responsive to branching to a subroutine. A subroutine is called from a calling module executing on a processor. Based on calling the subroutine, a value of a pointer to a reference data structure, such as a TOC, is predicted. The predicting is performed prior to executing a sequence of one or more instructions in the subroutine to compute the value. The value that is predicted is used to access the reference data structure to obtain a variable value for a variable of the subroutine.

Techniques for processing instructions include receiving a plurality of instructions from a program counter (PC) operable to be fused into a PC-relative plus offset instruction. The technique also includes fusing the plurality of instructions into an internal operation (IOP) that specifies PC-relative addressing with an offset. The technique also includes computing a shared PC portion that includes one or more common upper bits of a PC address of each of the plurality of instructions. If the shared PC portion is different than a previously computed shared PC portion, the technique transmits the shared PC portion to one or more downstream components in the processor pipeline. The technique further includes transmitting the IOP with a representation of lower order bits of the PC address and processing the IOP.

A digital signal processor having a CPU with a program counter register and, optionally, an event context stack pointer register for saving and restoring the event handler context when higher priority event preempts a lower priority event handler. The CPU is configured to use a minimized set of addressing modes that includes using the event context stack pointer register and program counter register to compute an address for storing data in memory. The CPU may also eliminate post-decrement, pre-increment and post-decrement addressing and rely only on post-increment addressing.

Apparatus and methods are disclosed for example computer processors that are based on a hybrid dataflow execution model. In particular embodiments, a processor core in a block-based processor comprises: one or more functional units configured to perform functions using one or more operands; an instruction window comprising buffers configured to store individual instructions for execution by the processor core, the instruction window including one or more operand buffers for an individual instruction configured to store operand values; a control unit configured to execute the instructions in the instruction window and control operation of the one or more functional units; and a broadcast value store comprising a plurality of buffers dedicated to storing broadcast values, each buffer of the broadcast value store being associated with a respective broadcast channel from among a plurality of available broadcast channels.

Apparatus and methods are disclosed for decoding targets from an instruction and transmitting data to those targets in accordance with a current instruction. Multimodal target hardware is used in conjunction with one or more of the routers so as to route data to an appropriate target. The data can be one or more operands or a predicate and the targets can include operand buffers, broadcast channels, and general registers. In this way, operands, for example, can be directed for use with multiple subsequent instructions, and there are multiple modes for distributing the operands to the multiple instructions.

Apparatus and methods are disclosed for controlling execution of memory access instructions in a block-based processor architecture using a hardware structure that generates a relative ordering of memory access instruction in an instruction block. In one example of the disclosed technology, a method of executing an instruction block having a plurality of memory load and/or memory store instructions includes decoding an instruction block encoding a plurality of memory access instructions and generating data indicating a relative order for executing the memory access instructions in the instruction block and scheduling operation of a portion of the instruction block based at least in part on the relative order data. In some examples, a store vector data register can store the generated relative ordering data for use in subsequent instances of the instruction block.

In an embodiment, a processor comprises a prefetch history array and a prefetch circuit. The prefetch history array comprises a plurality of entries corresponding to prefetch addresses, each entry of the plurality of entries comprising a sublength value associated with a frequency that a stride is repeated. The prefetch circuit is to: for each entry of the plurality of entries, adjust the sublength value based on stride matches for an address of the entry; adjust a short stream counter based on the sublength values of the plurality of entries in the prefetch history array; determine whether the short stream counter has exceeded a throttling threshold; and in response to a determination that the short stream counter has exceeded the throttling threshold, throttle a prefetch level of the prefetch circuit. Other embodiments are described and claimed.