The central processing unit, or CPU, is quite often referred to as the "brain" of the computer. Today's CPUs are like superbrains! A typical CPU today runs at 3GHz or higher, use two or more cores, and some can easily process 50 billion operations per second. That's a good deal more than we would have seen just 5 years ago. Some mornings I have trouble processing the thought need coffee! Of course we know that the human brain is much more sophisticated and functional than a CPU, but the CPU wins out when it comes to sheer calculating power.

You might hear the CPU referred to as a microprocessor, which technically it is. It's a much smaller version of the processors that were used 50 years ago. And although microprocessor might be a more accurate term, it has become more acceptable to refer to it as CPU, which this chapter does. However, you also see CPU manufacturers such as Intel refer to them as processors, so for all intents and purposes, the three terms mean the same thing. Keep in mind that a computer has other processors used by video cards and elsewhere, but know that the CPU is the main processor.

This chapter discusses some CPU technologies and cooling methods and talks about the models of CPUs offered by Intel and AMD. Afterward, the chapter demonstrates how to install and troubleshoot the CPU.

CPU 101
The CPU is often the most-expensive component in the computer; it's also one of, if not the, most important. The CPU's main function is to execute instructions or programs. Its speed, or clock rate, is measured in Hertz. For example, at 2.66GHz, a CPU operates at 2.66 billion cycles per second; we speak more to this concept in a moment. But although the speed of the CPU might be important, other factors should also play into your decision when choosing a CPU, including the chipset on the motherboard, CPU technology, and the brand of CPU. Chapter 2, "Motherboards," covers chipsets, but let's go ahead and talk about the various CPU technologies and brands of CPUs now.

CPU Technology
CPU technology is a key factor when considering a CPU. It all comes back to the motherboard; the CPU must be compatible with the motherboard in a number of ways. It is important to think about the speed (clock rate) of the CPU you want to use and whether that speed can be supported by the motherboard, and if the CPU fits in the motherboard's socket. Also, a decision has to be made as to whether to use a 32-bit or 64-bit CPU, and choose either a single-core or multi-core CPU; this will be based off the motherboard and the type of operating system you plan to install. Getting deeper into the technical side of the CPU, you might want to know the amount of cache included with the CPU, and the amount of power it requires.

Clock Rate
The clock rate is the frequency (or speed) of a component. It is rated in cycles per second and measured in hertz (Hz). For all practical purposes, the term clock rate is the same as the more commonly used term: clock speed.

Components are sold to consumers with a maximum clock rate, but they don't always run at that maximum number. To explain, let me use a car analogy. The CPU is often called the "engine" of the computer, like a car engine. Well, your car's speedometer might go up to 120MPH, but you'll probably never drive at that maximum—for a variety of reasons! When it comes to CPUs, the stated clock rate is the maximum clock rate, and the CPU usually runs at a speed less than that; in fact, it can run at any speed below the maximum.

Now, we're all familiar with speeds such as 2.4GHz, 3.0GHz, or 3.2GHz. But what is the basis of these speeds? Speed can be broken down into three categories that are interrelated:

•Motherboard clock speed: The base clock speed of the motherboard. Also referred to as the system bus speed, this speed is generated by a quartz oscillating crystal soldered directly to the motherboard. For example, the base clock speed on the motherboard used in Chapter 2 is 333MHz.
•External clock speed: This is the speed of the front side bus (FSB), which connects the CPU to the Memory Controller Hub (northbridge) on the motherboard. This is usually variable and depends on the CPU you install. In addition, it is determined from the base clock speed of the motherboard. For example, our motherboard's maximum external clock speed (or FSB) is 1333MHz. Simply put, this means that it is transferring four times the amount of data per cycle as compared to the original base clock speed. 333 MHz x 4 = 1,333MHz.
•Internal clock speed: This is the internal speed of the CPU. For this book I purchased the Intel Q8400 CPU that is rated at 2.66GHz. The CPU uses an internal multiplier that is also based off the motherboard base clock. The multiplier for this CPU is 8. The math is as follows: base clock speed x multiplier = internal clock speed. In our example, that would be 333MHz x 8 = 2.66GHz. Our motherboard can support faster CPUs also, for example, the Intel Q9650 that has an internal clock speed of 3.00GHz. This means that it has a multiplier of 9 (3.00GHz / 333MHz = 9). Some motherboards allow for overclocking (not ours), which enables the user to increase the multiplier within the BIOS, thereby increasing the internal clock speed of the CPU. This could possibly cause damage to the system, analogous to blowing the engine of a car when attempting to run a 10 second ¼ mile. So approach overclocking with caution.
NOTE
Quite often motherboard manufacturers state only the internal and external clock speeds (CPU and FSB); you might need to dig for more information concerning the base clock speed. To make matters more confusing, some manufacturers refer to the FSB as the system bus, but you can tell the difference. Just remember that the FSB is calculated from the base clock of the motherboard. Quite often, it's multiplied by four. Currently, FSBs are between 800MHz and 1600MHz.

However, the external clock speed (FSB) isn't actually a factor for AMD CPUs or newer Intel Core i7 CPUs because they have essentially done away with the FSB. Intel just recently started using the QuickPath Interconnect (QPI) technology in newer motherboards.

32-Bit Versus 64-Bit
The bulk of today's CPUs are 64-bit; it's a type of CPU architecture that incorporates registers that are 64 bits wide. These registers, or temporary storage areas, allow the CPU to work with and process 64-bit data types and provide support for up to one-terabyte of platform address space. 64-bit CPUs have been available for PCs since 2003. Examples of 64-bit CPUs include the AMD Phenom and Intel Duo Core CPUs.

The predecessor to the 64-bit CPU was the 32-bit CPU. Intel started developing well-known 32-bit CPUs as early as 1985 with the 386DX CPU (which ran at a whopping 33MHZ!), and AMD did likewise in 1991 with the Am386. A 32-bit CPU can't support nearly as much address space as a 64-bit CPU; 32-bit is limited to 4GB. Most editions of Windows are available in both 32-bit and 64-bit versions.

You will probably still see 32-bit technologies (such as the Pentium 4) in the field; however, due to applications' ever-increasing need for resources, these older CPUs continue to diminish, whereas 64-bit technologies (such as Core 2 Duo) will become more prevalent.

You might hear of the terms x86 and x64. x86 refers to older CPU names that ended in an 86—for example, the 80386 (shortened to just 386), 486, or 586 CPU and so on. Generally, when people use the term x86, they refer to 32-bit CPUs that enable 4GB of address space. x64 (or x86-64) refers to newer 64-bit CPUs that are a superset of the x86 architecture. This technology can run 64-bit software and 32-bit software and can address a maximum of 1TB.

Windows Vista and Windows XP come in 64-bit and 32-bit versions so that users from both generations of computers can run the software efficiently. Windows 2000 Professional was designed for 32-bit CPUs only.

This has been an excerpt from Chapter 3 "The CPU" of the A+ Exam Cram 5th Edition. Pick up your copy today!