Almost all microcontroller (and microprocessor) development systems use some form of a bootloader. Often called firmware, mistakenly, the Arduino bootloader is one example. Since it is a rather popular platform, let’s use it as an example. Let’s talk about what a bootloader does and how it works.

When a microcontroller turns on, it only knows how to do one thing.  Typically, that one thing is to run an instruction found at a specific memory location. Often this location address 0x0000, but not always. Usually, this memory location will contain a jump instruction to another place in memory, which is the start of the user program. The bootloader, however, exists in a slightly separate memory space from the user program.

On power-up or reset, a bootloader is a section of program memory that runs before the main code runs. It can be used to setup the microcontroller or provide limited ability to update the main program’s code.