A Comprehensive Review on Compiler in Neural Networks

It is significant that neural nets are at the origin of all computer science; with the expansion of technology, machine-learning, deep-learning, and artificial intelligence are now incorporated into machines and especially computers for the ease of human interaction. In this paper, we are interested in incorporating machine learning and neural networks with compiler design and showing how it leads to a reduction of time, optimization, and its implementation in the structure of the compiler. The primary purpose of this study is to establish a bridge between the compiler of a computer program and a neural network and proclaim different ways in which they are incorporated with each other. This summary paper will talk a lot about neuromorphic hardware, neural network compilers, or even an automatic RTL compiler as improving the scalability of the compiler. Again, we will also discuss neural code generation, the CNN accelerator, and more about feed-forward neural for predicting memory.