The term “bit” is short for “binary digit,” which can be a 1 or 0. In this context, the number of bits refers to how many 1’s and 0’s the computer’s processor can use to communicate instructions and assign memory addresses. When dealing with instructions, 64-bit software offers increased performance over 32-bit programs, because the processor will be able to deliver more information each time it issues a command. When dealing with memory address, a 32-bit string is limited to a little over 4 million bytes of memory, or 4GB. That means the computer cannot store data to RAM beyond 4GB, so installing more than 4GB on a computer that runs a 32-bit version of Windows is a waste.
A 64-bit system reaches a memory address limit at 18.45 exabytes—a number that’s around 4 billion times larger 4GB. It’s wise to go with a 64-bit operating system, as it’s backward compatible with almost any popular 32-bit application. Those with 64-bit versions of Windows, whether it be Vista, Windows 7, or Windows 8, should opt for 64-bit applications (if a choice is available), because the program will (all else being equal) run faster than a 32-bit counterpart. Note that if you run a 32-bit version of Windows 7 and opt for an Upgrade Assistant installation to Windows 8, you’ll end up with a 32-bit version of Win8. To “upgrade” to a 64-bit edition of Windows 8, you’ll need to buy a full version of Windows 8 and perform a Clean Install.