scope:

This standard defines a set of operator interfaces frequently found in artificial intelligence (AI) applications, where the AI operators refer to the standard building blocks and primitives for performing basic AI operations. It defines the functionality and the specific input and output operands of an AI operator. It considers both generality and efficiency and covers various types of operators, such as those related to basic mathematics, neural network, and machine learning. The defined operator interfaces serve as a reference for vendors to implement and use AI operators.

Purpose

This standard standardizes the development interfaces for commonly used AI operators, which helps to reduce the complexity of AI software and hardware adaptation, improve the efficiency and portability of AI software development, promote the evaluation and promotion of AI hardware, and promote the benign development of AI software and hardware ecology.