Programming terminology: Byte

Once you know what a bit means when it comes to programming, the byte is just one step away...

Byte structure

Basically, a byte - with regard to computer systems or microcontrollers and related digital signal processing - is just a group of eight bits. These bits directly follow each other with regard to their location within memory, cpu register, serial data etc.

Numerical representation

Similar to the bit, the byte is based on the binary system, as all eight bits can be zero or one in value (turned off or on).

Since binary numbers can get pretty long quickly (with each of the positions representing a value to the power of two), byte-values are typically represented with the help of the hexadecimal system. Since the hexadecimal system uses the numbers from zero to nine and the alphabet characters from "A" to "F", a single number or character can represent the decimal values from zero to 15 - which is equivalent to the what you can represent with eight binary numbers:

A single character hexadecimal value like in the above represents just one half of a byte - either the lower (values zero to fifteen) or the higher part (values sixteen and above). This "half-byte" is called a "nibble" - although this term is seldomly used these days (especially since data processing uses not just single bytes, but mega-, giga- or tera-bytes).

Just for completion, here is the table with the full byte values:

Byte processing

Early (personal and home) computers used 8-bit CPUs, so single byte values were used for most computing within the CPU. Since then, CPU technology (and the technologiy surrounding them) has progressed to 16-, 32- and currently 64-bit processor architectures. Thanks to this development, computers can handle larger numbers and address more memory, and thanks to faster processing speeds and parallel processing, data can be handled in bigger chunks ever faster. 

Still, blocks of processed data (or instructions) build up on top of bytes - just more of them at any one time.

And this is the reason why all the data-size related values for memory - either on a chip or on an external digital storage medium - still use bytes to describe the actual usable capacity:

• kilobyte: 1.024 bytes (since the base is binary, and each byte is made up of 8 bits, and 1.024 is 2 to the power of 10)
• megabyte: 1.024 kilobytes = 1.024 x 1.024 bytes = 1.048.576 bytes
  
• gigabyte: 1.024 megabytes = 1.024 x 1.024 kilobytes = 1.024 x 1.024 x 1.024 bytes = 1.073.741.824 bytes
• terabyte: 1.024 gigabytes (1.099.511.627.776 bytes)
  
• petabyte: 1.024 terabytes (2 to the power of 50 bytes = 1.125.899.906.842.624 bytes)
  
• exabyte: 1.024 petabytes

There is a nice article on the Lifewire website on the topic of computer-related data / storage sizes which you can find here. Naturally, they don't mention the nibble (half-byte) since this is not really used for direct computing (even not during 8-bit times when it was more useful).

Internal usage

One thing you have to remember - especially when you are working on more low-level stuff, maybe even in the Assembler programming language - is that byte-values are used to represent everything inside the computer: computing instructions, data that has to be processed, memory adresses etc.

Thankfully, hardware and BIOS create an ordered structure, so when the system is started, the computer knows where to start. The CPU instruction set - which is realized by the internal electronic circuits on the chip - then acts as a commando interpreter, which identifies the initial commands and executes them, processing following data values. This is all done on the hardware level, and forms the basis of every operating system and program.

Every program is - at the lowest level - just a big block of bytes, with commands and data all following each other in a format that the CPU can understand and execute.

Copyright © 2008 - 2020 Markus Dieterle, Alternative-Computer-Programming.com