¶ … Operating Systems Copernicus, Newton, Kepler, and Einstein all developed "laws" or theories so fundamental to our understanding of the universe that they are known to most people (Krass, 2003). Moore's notion that computing power would double about every 18 months as engineers figure out how to build ever-faster microprocessors is not as popular as Einstein's theories, but it has held up very well over the years.

Many of those advances, however, have been made within the confines of the 32-bit architecture. However, since the mid-'90s, some computers have been built with expanded memory and processing speeds, as a result of 64-bit architecture. While 64-bit processors have traditionally been expensive, proprietary, and complex computers used mainly by scientists and engineers, modern and relatively inexpensive 64-bit processors bring this speed advantage to the everyday processors that power Websites, corporate applications, and even personal computers.

Sixty-four-bit computing represents the third major architecture shift since the invention of the microprocessor. The first shift, in the early 1980s, took processors from 8-bit to 16-bit computing. The second came a few years later with the move from 16-bit to 32-bit computing. The third shift appeared in the mid-1990s, when the first proprietary 64-bit processors hit the market. More recently, Intel, Advanced Micro Devices, and Apple Computer have introduced 64-bit processors for desktop computers and servers running the Windows, Linux, and Mac operating systems.

Understanding the Differences Between 32-Bit and 64-Bit

It is important to understand what a 64-bit processor really does, and what potential advantages it has over a 32-bit processor (Dean, 2003). Many people believe that a 64-bit processor can work on twice as much information at the same time in comparison to a 32-bit processor. This is true in some ways, but does not equate to a doubling in performance, as many believe.

64-bit processor is simply one that can work with numbers that are up to 64 bits long instead of 32 bits long. Each bit is a binary digit represented by a 1 or a 0, and each additional bit to a binary number means it can represent a greater range of numbers. This range is known as the 'dynamic range'.

Increasing the dynamic range of a processor provides an advantage as it can more easily manipulate larger numbers. However, this is a limited advantage even in high-end applications. Most CPUs do not need to be throwing around numbers larger than 4.3 billion, and if it does come across one occasionally, it can always split it up, although at the cost of performance.

One of the true benefits of 64-bit processors is the ability to access more physical memory (Dean, 2003). At the moment, with complicated programming tricks and workarounds, a 32-bit processor can only access 4GB of memory, because each bit of memory must have an address so the processor can locate it. With 4GB of memory, a user has used up every number a 32-bit processor can represent as an address, and it is nearly impossible for the processor to be able to recognize more memory.

64-bit processor can potentially access up to 18 million terabytes (18 billion gigabytes). Not that any one person would want that much yet, but there are plenty of applications where 4GB is not enough, such as large database servers, 3D CAD and scientific analysis.

Performance benefits will increase when software is written specifically to take advantage of 64-bit processors. In this case, a 64-bit processor can perform an instruction on a larger chunk of data in one operation, but again, this is only of benefit if a user is working with very large numbers, or they need very high integer precision. As such, daily office applications and games realize very little benefit or improvement from 64-bit technology.

In a nutshell, the main performance-related benefit of 64-bit computing is the ability to access more memory. As many personal computers sold today are supplied with 512MB of RAM, it is reasonable to assume that PCs being shipped in future years time might come with 4GB as standard, in which case user will consider a 64-bit processor to increase their capabilities.

An additional benefit that comes from some 64-bit processors, including the 32-bit/64-bit AMD64 processors, is an increase in the number of registers, GPRs (General Purpose Registers) (Dean, 2003). Registers store each piece of data just before being operated upon, so the more registers, the more data can be made ready to process - although this raises another challenge of how best to manage these registers to make sure they're optimized. Still, more registers are an excellent thing for programmers.

How 64-Bit Processors Work

The term 64-bit describes the size of the addresses the processor uses to organize the system's main memory banks (Krass, 2003). Sixty-four-bit systems use wider registers so that the programs running on the computer can compute...

This expansion of the main memory address range greatly improves the overall ability of the chip to crunch data. While a 32-bit processor can directly address as many as 4 gigabytes (billion bytes) in the main memory, a 64-bit system can address as many as 16 exabytes or 16 billion gigabytes.
This means that computers with 64-bit processors can run database and other business programs faster, manage larger data files and databases, allow more concurrent users and applications to access data, and reduce software-licensing fees. Basically, the more memory a processor can access at one time, the less it relies on information stored on the computer's disk drive.

The less I have to go to disk, the faster my applications run," says Jeff Jones, director of strategy for IBM DB2 information-management software (Krass, 2003). "The faster my applications run, the quicker I can make business decisions."

Since memory costs have been lowered (a decade ago, 4 gigabytes of memory cost about $100,000; today the cost is under $2,000), the processor has become more of a bottleneck (Krass, 2003). The notion of a chip that can tackle 16 exabytes may seem like too much of a good thing, but companies are already figuring out ways to optimize their older applications to take advantage. New applications written expressly for 64-bit architectures are being developed by Microsoft and others.

64-bit processor, with 64-bit registers and a 64-bit integer data path, has the ability to address considerably more memory than a 32-bit processor, with 32-bit registers and a 32-bit integer data path. A 32-bit processor provides flat addressing of up to 232 32-bit bytes, or about 4GB of memory. On the other hand, a 64-bit processor provides flat addressing for 264 64-bit bytes, or 18 billion GB (18 exabytes) of memory. Scalability increases associated with 64 bits are dramatic and can be illustrated by the following example:

64-bit integer arithmetic

SOURCE: Krass, P. (November 17, 2003). Moore is merrier: for power users everywhere, your chip has come in. Hewlett-Packard Company. CTO.com.

In a 64-bit CPU, integer arithmetic is 64 bits providing greater performance and precision. Since most compilers support 64-bit data types, even on 32-bit CPUs, the main benefit for integer arithmetic is enhanced performance on larger data types.

The main benefit of 64-bit computing is increased scalability of your computer and applications. Some applications cannot fit into a 32-bit computing model. For example, limitations on file size in a 32-bit environment may require database systems to use multiple files to represent one single file. Applications requiring large files, a large number of files, or a large number of users benefit from 64-bit computing.

The following table summarizes the sources of increases in performance and scalability associated with 64-bit computing by type of application:

example sources of performance and scalability gains

Large database

Larger memory allocation per user

Many more users

Large file implementations

Reduced swapping

Decision support

Direct addressing

Reduced swapping

Large file implementations

Technical applications

Large process data space

More available shared memory segments

Reduced swapping

High-precision arithmetic

SOURCE: Krass, P. (November 17, 2003). Moore is merrier: for power users everywhere, your chip has come in. Hewlett-Packard Company. CTO.com.

Advantages and Disadvantages

64-bit computing has been in the public eye for more than a decade. In 1992, DEC (Digital Equipment Corporation - now a subsidiary of HP-Compaq) released the Alpha 64-bit processor, running at the then-incredible speed of 200 MHz (Dean, 2003). Since then, there have been a slew of 64-bit processors on the market, including the MIPS R. series, Sun UltraSPARC, IBM PPC 970 and more. These processors were all targeted at high-end computing applications, such as scientific analysis, or running large back-end servers.

However, in recent years, 64-bit computing has become more available to the masses, mainly in the form of AMD's new AMD64 architecture (previously known as x86-64), and, to a lesser extent, Intel's IA64-based processors. Both chip giants have taken very different approaches to the 64-bit proposition, and both are approaching it from different directions.

The main advantages of 64-bit are faster computing and lower IT costs, and these two benefits work simultaneously (Krass, 2003). The speed advantage means complex applications formerly requiring computers costing $25,000 and up can now be run on computers costing as little as $2,000.

Databases, which are used by nearly all businesses, should get a large boost. When a database accesses records on a drive, it leaves a copy…