Advantages of a 64 bit system

Question

From a developer perspective i am trying to understand , what is the selling point of a 64-bit system ?

I understand that more registers are at your disposal , more

Answer 1

Sorry, I kept this a bit list style, and avoided not mentioning increased memory addressing and more which had already been mentioned.

a) You'll have double the amount of registers, actually giving you three times the amount of registers at you disposal. b) General purpose registers are increased from 32 to 64 bit, meaning processing of 64 bit integers (i.e.) will be faster. c) More efficient position independent code since data can now be referenced relative to the RIP register. d) No-execute bit making the system more tolerant towards buffer overflows.

Answer 2

A 64-bit system has 64-bit memory addresses instead of 32-bit addresses, and so the maximum amount of memory available is 2⁶⁴ vs 2³². This first quantity is roughly 10¹⁸ bytes, compared to the 10⁹ bytes you can have with the latter quantity. Consequently, it's possible to have a lot more memory in a 64-bit system.

Answer 3

Most desktop apps actually don't need to be 64 bits binaries, to the notable exception of image or video processing apps. In contrast, server apps often do.

The best advantage of running on a 64Bits machine a is that your app will have more RAM space, so it won't get swapped away when other apps will need the RAM. Yet 32 bit apps run on 64bits machine (luckily enough). I'm typing this on a 16GB Linux laptop with 2 DB (Oracle 11g and Mysql 6.0alpha), Windows in VirtualBox and a host of Java VMs, eclipse with Xmx=2GB...) and I couldn't cram all this on 4GB. Yet I still prefer to run 32bits apps when they do not need large amount of memory space.

Answer 4

When you run multiple processes e.g. debug session, compiler and other tools you will notice a big performance gain if you have lots of RAM in your system. I have 16GB RAM in my Win7 system and I will never go back to having less. Its a bit like when you start using dual monitors, one just isn't enough after that.

Answer 5

As you said, more memory can be a big advantage. For 32bit systems you'll be limited to processes max 4GB (or even 2 or 3, depending how annoying your OS is).

64bits is double the amount of bytes per instruction, so you have more bandwith internally. eg: faster everything.

see also: http://lifehacker.com/5431284/the-lifehacker-guide-to-64+bit-vs-32+bit-operating-systems

Answer 6

A few Mac OS X specific answers (general ones are covered in other replies):

1) In 32 bit OSX, address space is mapped 4/4 (i.e. kernel gets the full 2^32 AND each app does), which requires flushing the TLB twice on each syscall. In 64 bit there's plenty of room to map the kernel and the application into different address ranges.

2) Objective-C programs use the new ABI/runtime on 64 bit x86 machines. This gets you C++ compatible exceptions, non-fragile instance variables, and some speedups.