We’ll focus on the two most significant improvements. First is expanded memory (RAM) capacity. With 32-bit CPUs an application can theoretically utilize a maximum of 4GB of RAM. By default, however, Windows only allocates 2GB for an application. (This can be increased to 3GB by setting a “large address aware” bit in an application.) 64-bit architecture extends this limit to as much as 1,024GB or 1 terabyte. Yes, that’s indeed much more. Access to such amounts of RAM allows you to load huge sample libraries or scores of audio tracks into memory, eliminating the throughput limitations of disk streaming and dramatically speeding up audio processing tasks such as pitch-shift or time-compression. (It will probably be some time before our motherboards and our wallets will be able to handle a full terabyte of RAM.)
The second advantage has to do with the amount of data that can be shoveled into a 64-bit CPU at one time. This is due to the increased number of registers available on x64 processors, as well as the improved design of the floating-point unit (FPU) on these processors. (Registers are the memory that resides closest to the CPU.) Data in a register can be accessed the instant the CPU needs it without any delay. 64-bit CPUs have twice the general-purpose registers and twice the FPU registers. Each general-purpose register is also twice as wide, 64 bits instead of 32 bits.