Combinational Circuits as Basic Building Blocks for AI

Overall Digital systems are categorized into two types of logic circuits:

• Combinational logic circuits.
• Sequential logic circuits.

When output depends upon the present state of inputs then it is known as combinational circuits. Logic gates (AOI i.e., AND, OR, NOT/INVERTER) are the basic building blocks to design different combinational circuits and for AI hardware they act as fundamental building blocks. AI accelerators such as GPUs and TPUs are used for deep learning in order to perform core AI tasks. Examples of Combinational circuits are adders and subtractors (Arithmetic circuits), multiplexers (MUX) and demultiplexers (DEMUX) (for data selection and distribution), magnitude comparator etc. For complex AI models the design of a combinational circuit depends on the type of circuit to be designed with specified inputs and outputs. Thus, different IC’s are used for more complex combinational circuits.