Calculation module and method

Abstract

The invention provides a calculation module. The calculation module comprises a memory, a register unit, a dependency relationship processing unit, a calculation unit and a control unit; the memory stores vectors; the register unit stores an expansion instruction; the control unit obtains the expansion instruction, analyzes the expansion instruction to obtain a first calculation instruction and asecond calculation instruction, determines an executive sequence of the first calculation instruction and the second calculation instruction according to the first calculation instruction and the second calculation instruction, and reads an input vector of the first calculation instruction from the memory; an index processing unit finishes operations related to index transformation, index-based data screening and the like; and the calculation unit finishes transformation of an index representation mode of input data of the first calculation instruction, screens the data according to indexes, and executes the first calculation instruction and the second calculation instruction for the input data of the first calculation instruction according to the executive sequence to obtain a result of the expansion instruction. The technical scheme provided by the calculation module has the advantages of low power consumption and low calculation expense.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to an operation module and a method. Background technique [0002] In modern general-purpose and special-purpose processors, more and more calculation instructions (such as vector instructions) are introduced to perform operations. Vector instructions are instructions that enable the processor to perform vector or matrix operations, such as addition and subtraction of vectors, inner product of vectors, matrix multiplication, and matrix convolution. In the vector instruction, at least one input is a vector or a matrix, or the operation result is a vector or a matrix. Vector instructions can perform parallel calculations by calling the vector processing components inside the processor to improve the operation speed. In the existing vector instructions, the vector or matrix in its operand or result is generally fixed in size. For example, the vector instruction in the vector e...

AI Technical Summary

Problems solved by technology

[0004] However, the existing vector operation instructions cannot support the operations of sparse vectors or sparse matrices in sparse neural networks, let alone the operations of variable-scale sparse vectors or sparse matrices, and the current vector operation instructions can only implement one type of operation. For example, one vector instruction can only implement one operation of multiplication and addition, but one vector instruction cannot implement more than two operations, so the existing vector operations have high computational overhead and high energy consumption.

Images