64-bit processors have registers that hold 64-bit values, double that of 32-bit processors. This translates to the ability to hold larger integers (numbers) in a single register, making operations involving massive numbers considerably faster. It should also lead to more precision in floating point numbers, except that current 32-bit processors already support up to 80-bit floating-point values in a single register. The major advantage of 64-bit computing is not in mathematical speed though, but in the amount of memory that a 64-bit processor is able to address, and consequentially use.

Memory addresses are run through the processor just like any other value, meaning they are stored in the registers. The largest integer number a 32-bit register can hold is around -2.1 to +2.1 billion. This translates to a maximum of 4GB of physical memory.

Various workarounds have been invented for the server market to transcend this limitation, but all sacrifice performance. 64-bit registers can effectively address up to 16 terabytes of physical memory which, to paraphrase Bill Gates, ought to be enough for anybody.

It should be noted at this point that 64-bit computing is specific to the operation of the processor and the software that is feeding it, not the other computer hardware in a system. There are no special '64-bit' computer memory chips or other peripherals. The AMD Athlon 64 processor uses the same PC hardware as conventional 32-bit systems.

AMD hit a home run in April of 2003 when it introduced its first line of 64-bit processors. The Opteron line was significant in that it allowed businesses the chance to upgrade towards the future of Windows (or the present day of Linux) without having to rejig their entire setup. Up until that point, the only 64-bit solution was Intel's expensive Itanium processor. The Itanium was powerful, but could not run existing 32-bit code natively. This meant businesses either had to adapt their software to work with the 64-bit Itanium and a 64-bit operating system, or purchase new applications that could. Needless to say, neither option is efficient or affordable.

The AMD Opteron was revolutionary in that it supported the traditional x86 architecture and 32-bit code that PC processors have been chewing on for years, in addition to software written specifically for 64-bit processors. This allowed businesses the luxury of upgrading their existing server hardware without the attending compatibility worries.

AMD64 - AMD's No Mess Solution for 64-bit

AMD achieved this by extending the x86 architecture to support their 64-bit processors, calling the resulting hybrid 'AMD64.' The AMD64 architecture allows for processors with two completely separate modes of operation.