Bus Width

You might have heard the term “bus” in connection with computers. A bus is a connection or channel between the chip set on your motherboard and the different parts of your PC, such as the processor, memory, expansion cards, etc.

There is the processor bus which connects the chip set to the CPU, the memory bus which connects the chip set to the memory, local buses such as the PCI bus (Peripheral Component Interconnect), expansion buses such as the ISA bus (Industry Standard Architecture), and more.

Each bus consists of two parts, the data bus, which transfers the actual data, and the address bus, which transfers information about where the data is supposed to go to.

Every bus runs at a certain clock speed, which is expressed in MegaHertz (MHz). For example, the old ISA bus runs at a whopping 8 MHz, the PCI bus runs at 33 or 66 MHz, and the AGP bus runs at 66 MHz.

Each bus also has a certain width which brings us finally back to our topic, because the bus width is expressed in bits, indicating how many bits can be transferred simultaneously over the bus. Physically, this means how many wires are in the bus. The ISA bus width ran from 8 to 32 bit, the PCI bus is either 32 or 64 bits wide, and the AGP bus is 32 bits wide. Naturally, this means the wider the bus, the more data can be transferred at the same time, the faster the bus is.

FAT 16 vs. FAT 32

When you partition a new, clean hard drive, one of the things you do is decide what file system to use. Until a few years ago, FAT16 was the main file system for home PCs. It’s main limitation was that it could handle only partitions up to a size of ca. 2 GB. The reason for this was that it used a 16-bit number to index each cluster.

Let’s do some math to understand the difference between FAT (or more exact FAT 16) and FAT 32. When you partition and format the drive, it is being organized into a lot of cubby holes, so to speak. Those are called clusters, which are numbered by the FAT to keep track of them and what’s in them. FAT 16 uses a 16 bit number to number them, and the highest number you can display using 16 bits, is 2^16 = 65,535.

The biggest possible size each cluster can be is 32,768 Bytes. Therefore, if you take the maximum number of clusters – 65,535 – times the maximum cluster size – 32,768 Bytes -, you get the maximum hard drive size FAT can handle – 2,147,450,880 Bytes, or 2 GB.

In contrast to that, FAT 32 uses a 32-bit number, and therefore can handle up to 2 TeraBytes