Convolution operation structure and method thereof

Abstract

The invention provides a convolution operation structure and a method thereof. The convolution operation structure comprises a conversion unit structure, the conversion unit structure comprises a first register group, a second register group, a multiplier, an adder and an intermediate result register group, wherein the first register set and the second register set are controlled to be in the formof a double-register set; in the first round calculation process, matrix multiplication is executed on a first matrix and a second matrix which are stored by the first round calculation process and the second round calculation process through the multiplier and the adder, and a first result matrix obtained through matrix multiplication is stored in the intermediate result register set; and the second register group and the intermediate result register group are controlled so as to register the intermediate result. In the second round calculation process, matrix multiplication is performed ona third matrix and the first result matrix which are respectively stored by the third matrix and the first result matrix through the multiplier and the adder, and a second result matrix obtained by the matrix multiplication is stored in the first register group.

Description

technical field [0001] The invention relates to the improvement of the neural network processor including the convolution operation unit, in particular to the improvement of the convolution operation structure capable of realizing the winograd convolution operation. Background technique [0002] Deep learning technology has developed rapidly in recent years, and it has been widely used in solving advanced abstract cognitive problems, such as image recognition, speech recognition, natural language understanding, weather prediction, gene expression, content recommendation and intelligent robots. And has excellent performance, so it has become a research hotspot in academia and industry. Deep neural network is one of the perception models with the highest level of development in the field of artificial intelligence. This type of network simulates the neural connection structure of the human brain by building a model, and describes the data features layered through multiple tran...

AI Technical Summary

Problems solved by technology

However, the existing winograd convolution operation unit is used to realize the above-mentioned matrix conversion hardware structure and the matching processing operations are very complicated, resulting in relatively poor circuit area and utility of the winograd convolution operation unit

Moreover, since the convolution operation of winograd is not ideal for the calculation effect of certain network sizes (for example, the acceleration effect of the convolution operation with a convolution kernel size smaller than 3×3 is not obvious, and the calculation speed is even lower than the traditional convolution operation Unit), so many neural network processors equipped with winograd convolution operation unit will also set up traditional convolution operation unit to deal with the calculation requirements for different network sizes

In these neural network processors, there is no computing unit that can be multiplexed for traditional convolution operations and winograd convolution operations, which makes the neural network processors need to be set up for traditional convolution operations and winograd convolution operations. corresponding circuit

Images