The ABCs of Semiconductor Technology
The world of semiconductors is rife with terms and acronyms whose meanings aren’t always intuitive. And each new wave of technology innovation brings with it another set of expressions to be learned and absorbed. Understanding these acronyms is crucial if you work with semiconductor-based technologies.
Even if you don’t, you may want to learn a bit about them given how much chips have come to permeate the news cycle – from legislation designed to pave the way for development of new semiconductor technology and more secure supply chains, to construction of new chip production facilities, to massive chip content in cars and other everyday items.
This post defines and demystifies a few select semiconductor industry acronyms so that you can use (or at least read about) them with more confidence. It’s by no means comprehensive, but it’s a good start.
Although it’s not an acronym, let’s start with why a chip is called a semiconductor. The chip is built on silicon, which is neither a conductor nor an insulator, but has properties in between, so the chip is a “semi-conductor.”
Compound semiconductors – different types of chips that combine two or more elements – offer attractive properties for power, communications, photonics and other applications. Some of the most common are silicon carbide (SiC), gallium nitride (GaN), and gallium arsenide (GaAs).
IC – Integrated circuit
ICs (i.e., chips, microchips or semiconductors) are miniature electronic circuits fabricated on a semiconductor material. They typically contain millions to billions of transistors, resistors, capacitors, as well as other electronic components.
CPU – Central processing unit
The brain of a computer system, the CPU is an IC that carries out instructions, performs calculations, and manages data flow. A logic device, it carries out such tasks as running programs, handling input/output operations, and managing memory. A microprocessor (MPU) is like a CPU in terms of computing power, but instead of being general-purpose, it’s purpose-built for a specific task – all CPUs are MPUs, but not all MPUs are CPUs.
MCU – Microcontroller
Created to control electronic devices, a MCU is a type of IC dedicated to performing a specific task for a designated application. The heart of an embedded system, an MCU comprises a CPU and various types of memories. They are less complex than MPUs, with fewer instructions to process.
GPU – Graphical processing unit
GPUs are specialized processors designed to speed graphics rendering and deliver images and video for viewing on smartphones, gaming systems, PCs, HDTVs and other electronic displays. GPUs must be able to handle large volumes of calculations quickly to yield high-quality results.
ASIC – Application-specific IC
An ASIC, as its name suggests, is an IC either fully or partially custom-designed for a specific use or application, unlike general-purpose chips such as CPUs or MCUs. An ASIC combines several different circuits all on one chip, making it a type of system-on-chip, or SOC. Uses for ASICs vary widely, but they basically involve controlling the functionality of an electronic device.
FPGA – Field-programmable gate array
A type of programmable logic device (PLD), FPGAs are multiple-purpose microchips that include circuitry enabling them to be reprogrammed by a system developer for use in multiple applications. Typically, FPGAs are meant for more complex designs while PLDs are used for simpler ones.
DSP – Digital signal processor
DSPs are processors designed to perform high volumes of mathematical functions without consuming a lot of power. They take analog voice, audio, video and other types of signals and quickly performing number crunching to convert them to signals we can interpret. Smartphones, smart TVs and wireless headphones are examples of devices that integrate DSPs.
MEMS – Microelectromechanical system
A MEMS is a miniature machine that combines both mechanical and electronic components to create sensors. Some examples of MEMS devices include accelerometers for airbag sensors, microphones, blood pressure sensors, optical switches, tire pressure sensors, biosensors, and thermostats – to name just a few.
MOSFET – Metal-oxide field-effect transistor
Electronic devices used to switch or amplify signals in a circuit, MOSFETs are used widely in communications and industrial and applications. Power MOSFETs are employed in automotive electronics, e.g., as switching devices in electronic control units or as power converters in EVs.
RAM – Random-access memory | ROM – Read-only memory
RAM and ROM are the two basic types of semiconductor memories, with many variations. Computers store the data they process in memory storage units, which rely on memory chips. The basic difference between the two is that RAM is temporary (or read-write), so data stored in it is written over with new information, while ROM is semi-permanent and used to retrieve pre-determined data or settings on a computer or other electronic device that can’t be written over.
Common versions of RAM are dynamic RAM (DRAM), typically used as a computer’s main memory, which must be regularly refreshed to retain the data, and static RAM (SRAM), which holds data indefinitely without refreshing. [For more in-depth information on the various types of memories and their applications, check out TechTarget.]
PCB – Printed circuit board
PCBs are non-conductive materials to which completed chips (sometimes packaged, sometimes not) can be mounted and electrically connected for integration into an end product. The photo shows a typical example of a PCB with a packaged IC and electronic traces.
KGD – Known good die
Individual chips are called die when manufactured on a wafer. Die that have been tested using specialized automated test equipment (ATE) and proven functional are called KGDs. Bad die are removed from the manufacturing line to ensure that they don’t end up in an electronic device.
EDA – Electronic design automation
Software and methodologies for designing chips, boards and systems are collectively known as EDA. As with chip manufacturing, which involves many different machines, each of which performs a specific function to create the completed IC, EDA involves many different steps. Because chipmaking has become increasingly complex, chip designers must take the testability and manufacturability of their design into consideration. EDA providers work closely with equipment makers to help integrate design for manufacturing (DFM) and design for test (DFT) techniques into their tools.
AP – Advanced packaging
Semiconductor packaging refers to the process of enclosing or encapsulating ICs and other electronic components to protect them from physical damage, provide electrical connections, and facilitate their integration into electronic systems. Some examples of package technologies include 2.5D/3D (which refers to the internal vertical structure of a package), flip chip, wire bonding, and chiplets. A system-in-package (SiP) contains several ICs, including a MPU, on a single substrate such as ceramic or laminate. [For all things packaging related, visit 3D InCites.]
As noted earlier, there are far more semiconductor industry acronyms out there than can be addressed in a single blog post. Moreover, different parts of the ecosystem, as well as targeted end markets – from automotive to lighting to communications – each have their own terminology and associated acronyms.
Hopefully, this piece has provided a useful primer on the alphabet soup of semiconductor industry acronyms. Be sure to check out the links provided above if you’d like to learn more on those topics.
Additionally, many industry organizations and players offer glossaries of industry terms on their own websites. Here are a couple of good ones: