What Is Firmware? Definition, Types, and Examples
Intermediate – Web/Development
What is firmware? Firmware is software written into a device’s non-volatile memory that controls, monitors, and manages the hardware it lives on. It sits between the physical components and any higher-level software, and it runs before an operating system does. Every laptop, router, printer, SSD, smart TV, and engine control unit has firmware inside it.
Firmware is usually written in C or assembly, then compiled into machine code for that specific device’s processor. That is where the “low-level" description comes from. The code is not easier for hardware to read than any other compiled program. It is simply written close to the metal, with direct access to registers, memory addresses, and peripherals that ordinary applications never touch.
What Is Firmware vs. Software vs. a Driver?
The three get used interchangeably, and they are not the same thing. The clearest way to separate them is by asking which processor runs the code.
| Firmware | Device driver | Application software | |
|---|---|---|---|
| Where it lives | Non-volatile memory on the device itself (ROM, EEPROM, flash) | The host operating system, on the host’s storage drive | The host operating system, installed by the user |
| Which processor runs it | The device’s own controller or microcontroller | The host CPU, in kernel space | The host CPU, in user space |
| What it talks to | The hardware directly | The OS on one side, the firmware or hardware on the other | The OS, through APIs |
| Who updates it | The manufacturer, through a flashing process | The OS vendor or hardware vendor, through driver packages | The software publisher, through normal updates |
| Examples | UEFI/BIOS, SSD controller firmware, router firmware | Nvidia display driver, printer driver | Excel®, Chrome, QuickBooks |
A printer makes this concrete. The printer’s firmware runs on a chip inside the printer and decides how the rollers turn and how much ink to lay down. The printer driver runs on your computer and translates a document into instructions the printer understands. Neither one is the other.
Firmware runs on the device’s own processor. A driver runs on the host CPU and tells the host how to talk to that device.
Where Firmware Is Stored
Firmware lives in non-volatile memory, which keeps its contents when power is removed. This is the opposite of random access memory (RAM). RAM is volatile: cut the power and everything in it is gone. That distinction is the whole reason firmware is stored the way it is.
Common non-volatile storage types for firmware:
- Mask ROM, written during chip fabrication and never rewritable
- PROM and EPROM, programmable once or erasable with ultraviolet light
- EEPROM, erasable and rewritable electrically
- Flash memory, the standard today, rewritable in blocks and fast enough for field updates
Storing firmware in non-volatile memory does not mean a device resumes an interrupted task after a power cut. It means the device still knows how to start itself when power returns. A washing machine that loses power mid-cycle has not saved your cycle. Its firmware is simply still there, ready to boot the control panel again.
Flash memory changed the economics of this. Once firmware could be rewritten in place, manufacturers could ship fixes after a product left the factory, and the annual firmware update became a normal part of owning hardware.
Types of Firmware
Firmware is generally sorted into three tiers based on how deeply it is embedded and whether it can be changed.
| Type | Stored in | Updatable? | Typical use |
|---|---|---|---|
| Low-level firmware | Mask ROM or one-time programmable memory | No, it is written during manufacturing | Basic hardware initialization on simple chips |
| High-level firmware | Flash memory or EEPROM | Yes, through a flashing or update process | UEFI/BIOS, routers, phones, SSDs, smart appliances |
| Subsystem firmware | Memory inside a component or subsystem | Sometimes, often bundled with a parent update | Baseboard management controllers, LCD units, storage controllers, embedded CPU microcode |
The phrase “internal firmware" gets used loosely to mean firmware that cannot be updated in the field. Low-level firmware in mask ROM genuinely cannot be rewritten, and replacing it would mean replacing the chip. Most modern devices do not work this way. Phones, cars, routers, and PCs almost all ship with high-level firmware in flash memory precisely so it can be patched.
Not all firmware is permanent. Low-level firmware in mask ROM cannot be rewritten. High-level firmware in flash is designed to be updated.
What Is Firmware Used For? Everyday and IT Examples
Everyday appliances are the easiest starting point. A washing machine’s firmware reads the cycle you select on the panel, then signals the motor, valves, and heater to run in sequence and produce the load you asked for. A microwave’s firmware handles the timer, the magnetron, and the turntable. A TV remote’s firmware reads which button you pressed and emits the matching infrared code.
The examples that matter more in an IT context sit deeper:
- UEFI and BIOS. The firmware that initializes a PC’s hardware, runs power-on self-test, and hands control to the operating system. Unified Extensible Firmware Interface (UEFI) has largely replaced the older Basic Input/Output System (BIOS).
- SSD and HDD controller firmware. Manages wear leveling, error correction, and how data is physically written.
- Router and switch firmware. Handles routing tables, NAT, firewall rules, and the admin interface.
- Baseboard management controllers (BMC). Server-side firmware that allows out-of-band management even when the server’s OS is down.
- Engine control units (ECU). Automotive firmware governing fuel injection, timing, and emissions, increasingly updated over the air.
- Peripheral firmware. Keyboards, webcams, monitors, and docking stations all carry their own.
Why Firmware Updates Matter
Vendors ship firmware updates for three reasons: to fix bugs, to add or change features, and to close security holes. The third one is the reason firmware patching belongs on a maintenance schedule rather than in a drawer of things to get around to.
What is firmware doing that turns an unpatched device into a security problem? It executes before the operating system loads, which puts it below every security tool that runs inside the OS.
Malware that establishes itself at the firmware layer survives an OS reinstall and a disk wipe, and endpoint detection tools running in the OS have limited visibility into it. In 2025, researchers at Binarly disclosed four high-severity UEFI vulnerabilities affecting more than 240 Gigabyte motherboard models, each carrying a CVSS score of 8.2 and each capable of hosting an implant that persists through a full reinstall (CERT/CC advisory VU#746790, 2025). ESET researchers reported a related class of Secure Boot bypass issues to CERT/CC in February 2026, with the vulnerable applications revoked in Microsoft®’s June 2026 update cycle (ESET Research, 2026).
The National Institute of Standards and Technology publishes the reference framework for this problem in SP 800-193, Platform Firmware Resiliency Guidelines, which sets out protection, detection, and recovery requirements for platform firmware. NIST SP 800-147 covers BIOS protection specifically.
Firmware updates carry their own risk, which is worth stating plainly. An interrupted flash can leave a device unbootable, so updates should be applied with stable power and with the vendor’s own tooling.
Firmware sits below the operating system. Malware that reaches it survives an OS reinstall.
Firmware in Cloud Desktop Environments
Firmware management is one of the quieter costs of running physical endpoints. Every laptop and desktop in a fleet carries UEFI, storage controller, and peripheral firmware, and each of those has its own patch cadence, its own vendor tooling, and its own failure modes.
Desktop virtualization changes where that work happens without making it disappear. When desktops run as Desktop as a Service, the operating system and applications execute on server hardware in our global data centers, and we handle firmware and patching on the host platform underneath. Our 99.95% uptime SLA and platform features cover that layer. Customers running Windows® cloud servers or a Linux VM with us do not flash BIOS on the machines their desktops actually run on.
What does not disappear is the endpoint. A thin client, a Chromebook, or a personal laptop connecting to a cloud desktop still runs firmware, and that firmware still needs patching. This matters most in BYOD and remote work setups, where IT has the least control over the hardware in use. The mitigation is that the endpoint holds no company data. Nothing is stored locally, so a compromised or replaced device is a hardware problem rather than a data breach, and our security and compliance controls plus secure remote access sit around the session rather than the device.
For teams without a dedicated patching function, V2CloudCare covers managed maintenance, and IT managers running lean can see how this shifts the workload. MSPs managing firmware across multiple client fleets tend to feel the difference first.
Firmware Frequently Asked Questions
What is firmware in simple terms?
It is the built-in software that tells a piece of hardware how to behave. Your microwave has no operating system and no apps, but something still has to read the keypad and run the timer. That something is its firmware.
What is firmware in a computer?
In a PC, the most visible firmware is UEFI or BIOS, which initializes the hardware at startup and hands control to Windows or Linux. A computer holds far more than one piece of firmware though: the SSD controller, the network card, the webcam, the keyboard, and the docking station each carry their own.
Is firmware hardware or software?
Software. Firmware is program code, not a physical component. It gets called firmware because it is stored on the hardware itself and is not casually installed or removed the way an application is.
What is the difference between firmware and an operating system?
Firmware initializes and controls specific hardware and runs first. An operating system loads afterward, manages resources across the whole machine, and provides an environment for applications. On simple devices like a microwave, the firmware is the only software present and no OS exists.
Can firmware be updated?
Most firmware can. High-level firmware stored in flash memory or EEPROM is designed to be updated, which covers phones, routers, PCs, SSDs, and modern vehicles. Low-level firmware written to mask ROM during manufacturing cannot be rewritten.
What happens if a firmware update fails?
The device can be left in an unbootable state, often called bricking. Many devices now ship with a backup firmware image or a recovery mode to survive this. Use the manufacturer’s official update tool and keep the device powered throughout.
Is BIOS firmware?
Yes. BIOS (Basic Input/Output System) is the firmware that initializes PC hardware at startup. UEFI (Unified Extensible Firmware Interface) is its modern replacement and supports larger drives, faster boot, and Secure Boot.
Where is firmware stored?
In non-volatile memory on the device: mask ROM, PROM, EPROM, EEPROM, or flash memory. Flash is the standard on current hardware because it can be rewritten, which is what makes field updates possible.
Keep learning: What is RAM? | Desktop as a Service | Secure remote access | Our data security | See pricing plans | Talk to a cloud expert
Get Started Today With V2 Cloud!
