What Are CMOS Integrated Circuits?

CMOS integrated circuits are chips that implement electronic functions by combining complementary pairs of n-channel and p-channel metal-oxide-semiconductor field-effect transistors (CMOSFETs) on a common silicon substrate. The complementary pairing minimizes static power dissipation because one transistor in each pair is always in a blocking state during a stable logic level, a property that made CMOS the dominant platform for processors, memory, radio transceivers, and sensor interfaces. The field spans digital, analog, mixed-signal, and radio-frequency circuit design, all sharing the same fundamental device physics while differing in the performance metrics that govern design choices.

CMOS integrated circuit technology has followed decades of miniaturization, with the minimum transistor gate length shrinking from tens of micrometers in the early 1980s to a few nanometers in current process nodes. Each generation of scaling expands the functions that can be integrated on a single die and reduces the cost per transistor, driving both the consolidation of system functions and the emergence of new application domains.

CMOS Device Fundamentals

The MOSFET is the unit building block of all CMOS circuits. An n-channel MOSFET conducts when its gate voltage exceeds the threshold voltage, pulling a node toward the negative supply; a p-channel MOSFET conducts when its gate is pulled below threshold, pulling toward the positive supply. Together, a complementary pair forms an inverter that draws current only during the switching transition. As process nodes scale below 10 nm, short-channel effects, including drain-induced barrier lowering, require multi-gate architectures such as FinFETs and gate-all-around nanosheets to maintain electrostatic control of the channel. These structural changes preserve the complementary switching principle at dimensions where planar gate structures would leak excessively.

Mixed-Signal Integrated Circuits

Mixed-signal integrated circuits combine analog and digital functions on the same die, bridging the physical world and digital computation. Analog-to-digital converters, digital-to-analog converters, phase-locked loops, and bandgap voltage references coexist with digital logic blocks in system-on-chip designs for smartphones, automotive controllers, and wireless sensor nodes. A persistent challenge in mixed-signal design is substrate noise coupling: fast-switching digital logic injects noise into the shared substrate, degrading the performance of sensitive analog circuits nearby. Guard rings, deep n-well isolation, and differential signal architectures are standard mitigation strategies. Mixed-signal neuromorphic computing circuits using hybrid CMOS-RRAM integration illustrates how mixed-signal design extends to emerging applications that blend conventional CMOS with novel memory devices for in-memory computation.

Radio Frequency Integrated Circuits

Radio frequency integrated circuits (RFICs) implement the transmit and receive signal chains in wireless communication systems using standard CMOS or BiCMOS processes. Key building blocks include low-noise amplifiers (LNAs), which set the noise figure of a receiver front end; mixers, which translate the signal between radio and baseband frequencies; voltage-controlled oscillators for carrier synthesis; and power amplifiers for the transmit path. Design of RF-CMOS integrated circuits for wireless communications describes how silicon CMOS processes have matured to enable single-chip transceiver integration that was previously feasible only in compound semiconductor technologies. Millimeter-wave CMOS RFICs now support 5G, 60-GHz Wi-Fi, and automotive radar applications, as documented in millimeter-wave CMOS RFIC development and applications, exploiting the transit frequency gains available from sub-10-nm nodes.

Applications

CMOS integrated circuits have applications in a wide range of fields, including:

• Microprocessors, graphics processors, and system-on-chip designs in computing and mobile devices
• SRAM and DRAM memory arrays in computers, servers, and embedded systems
• Radio transceiver chips in cellular, Wi-Fi, Bluetooth, and radar systems
• Neuromorphic processors that implement spiking neural network models for edge AI inference
• Sensor interface and data acquisition chips in industrial and biomedical instrumentation
• Power management and battery charging integrated circuits in portable electronics