What's the real deal with 64-bit computing in Snow Leopard?

28.08.2009
One of the biggest points of confusion around Apple's newest version of Mac OS X is about whether it's really a 64-bit or a 32-bit operating system. Apple bills as supporting 64-bit from top to bottom, while some industry watchers say it's not a true 64-bit OS. What gives?

Both Apple and Microsoft have been chasing 64-bit computing as if it were the Holy Grail -- and in some ways it is. 64-bit computing opens a lot of doors for hardware and software developers. Because 64-bit processors have twice the number of registers to work with, they can process twice as much information per clock cycle as 32-bit processors. That obviously means much better performance.

Another big advantage to 64-bit computing is addressing memory. Applications have traditionally been written to address the memory space available to 32-bit processing, which tops out at 4GB. If a system has more RAM than that, any applications that are designed to work with 32-bit processing will not be able to interact with any memory beyond 4GB.

By contrast, applications written for 64-bit operation can address and store data in up to 16 exabytes -- that's 16 billion GB -- of RAM. For memory-hungry applications that perform complex and data-intensive tasks (such as high-end graphics, video, or scientific computing tools), this alone can make a huge difference in performance, regardless of the number of registers that the processor has to offer.

In fact, 64-bit computing even offers some additional security because the routines used to interact with the processor are more secure than in 32-bit computing. What's more, the system heap (which is shared memory available to applications) is designed with both hard protections and stronger checksum algorithms that help protect against attempts to corrupt the addressable memory used by the operating system and applications.