Kurt Shintaku's Blog

I’ve been asked several times, "Why should I use 64-bit processors?  In English please."

So here’s a quick list of reasons, followed by an old writeup that goes into this more deeply.  There are 3 reasons – big ones – that 64-bit processing is the future:

1. GREATER MEMORY
   Windows XP Professional x64 Edition can address 128 GB of RAM.  Windows XP Pro 32-bit can address 4 GB RAM.  Manipulating data in memory is 100+ times faster than trying to do the same thing on hard drives and this means that, assuming you have the RAM, you could do a lot more in memory using 64-bit CPUs than you could using 32-bit CPUs.
2. FASTER GENERAL COMPUTATION
   64-bit CPUs can store twice as many bits in its registers allowing them to "number crunch" at least twice as much information within the processor at the same time than 32-bit CPUs.  In addition, the chip itself has data I/O throughput that is twice as much as 32-bit CPUs.  This means that not only can 64-bit CPUs access more memory, even within the processor itself, the chip is optimized to only have to retrieve data from memory HALF as often as 32-bit CPUs to accomplish the same tasks.
3. HIGHLY ACCELERATED EXECUTION FOR SELECT OPERATIONS
   64-bit CPUs expose a larger instruction set to operating systems than 32-bit CPUs.  Instead of running multiple CPU cycles with base-level instructions to accomplish the same task in software, hardware instructions allow the same computational task to be completed in a single cycle.  Most every CPU generation produced expands the instruction set available to operating systems, however it’s important that the OS actually take advantage of those instructions to realize any performance improvement and that’s where 64-bit operating systems like Windows Vista x64 & Windows Server R2 x64 come into play.

———————–

<excerpt from Microsoft’s online description of 64-bit computing>

The 64-bit versions of the Windows Server 2003 family are designed to run on high-performance 64-bit processors. These processors improve performance and scalability by processing more data per clock cycle, addressing more memory, and running faster numeric calculations.

For workloads that are memory-intensive or compute-intensive, the performance and scalability benefits are often dramatic. Large data sets can be loaded entirely into memory, reducing the need for slower disk access; complex calculations that take hours to complete on 32-bit systems can be done in minutes; workloads that once required large server farms can be deployed on a single server.

Ultimately, these breakthroughs lower your total cost of ownership (TCO) by delivering superior price performance and reducing administration costs. At the same time, the 64-bit versions of Windows Server 2003 share the same familiar user interface and development environment as their 32-bit counterparts, allowing you to leverage your existing Windows skills. Because 64-bit Windows Server 2003 systems integrate seamlessly with other members of the Windows Server family, you can add 64-bit systems for the workloads that need it, while keeping the rest of your 32-bit environment intact.

The Windows Server 2003 family supports two different 64-bit architectures. The first is called Explicitly Parallel Instruction Computing (EPIC), and supports the Intel Itanium processor family. This version is well-suited for native 64-bit applications requiring the highest levels of scalability.

The second 64-bit architecture supported by the Windows Server 2003 family is based on 64-bit extensions to the x86 instruction set, and supports both AMD64 processors and Intel processors with Extended Memory 64 Technology (EM64T). This platform lets you take advantage of the wealth of existing 32-bit Windows applications, while also powering the latest 64-bit technology.