How A Central Processing Unit Works In Personal Computers

Principles Of The Central Processing Unit

Although most processors vary in cosmetic features such as size and shape their underlying principles remain the same. The CPU is the brains of any computer system; it processes data and completes number crunching and formula calculations based on a set of rules and instructions defined by software or hardware. These applications can be operating system components, device drivers, and executable or program files. The processor is there to decipher what needs to be done, and passes the processed information back to the device that called on the CPU to complete the task.

CPU Languages

Like all peripheral devices inside a computer, the CPU communicates with other devices in a PC using a binary language system, a digital language that only a computer can understand.



Also referred to as the binary numbering system, this type of language is simply a mixture of 1's and 0's corresponding to a human value that represents a particular instruction or set of instructions.

The processor knows exactly what to do with these combinations because that is how it's programmed. You can think of the binary numbering system as a light bulb. Data flows through a computer in random sequences of 1's and 0''s. These 1's and 0's are actually electrical pulses similar to a light bulb being powered on and off from it's associated switch as illustrated in the diagram shown below.



One example to further explain the concepts described above include how you actually see output on your computer's monitor. Before any text or images appear on your screen they must pass through a number of devices before the data is actually put into a human format, or viewable on the monitor.

The video card passes a signal from a particular software application or hardware component through to the operating system at which time the operating systems hands over control of the instruction to the processor.

The processor thinks about what it should do with the data and completes any calculations requested by the operating system. Once the data is processed it is sent back to the video card. It's at this point the video system kicks in again and finishes processing the data. The end result, is the function that you asked your particular software application to complete has been completed on the screen, and the result is referred to as output.

In this example, the processor is the most integral part of the process; the video card, computer monitor, and other devices are there simply as extra variables for the processor to complete the task. It's the actual processor that completes that tasks of thinking, calculating, and transferring of data.

Transistors In A Processor

The electrical transistors located on the bottom of a processor chip are the apparatus that can interpret magnetic pulses (or data that is passed through to other components of the processor).

When data travels through transistors numbers are crunched, formulas are processed, and data is either inputted or outputted to other components in the system.

The actual number of transistors on a given processor varies depending on the type of processor being used.

After other parts of the processor are finished their jobs, the transistors facilitate the communication of data back to other devices in the computer. One thing to remember about this step is that all this number crunching going on is happening virtually instantaneously through the chip's circuitry. Of course, the slower the processor the less transistors it will have, and the slower it will perform.

Layers Of Memory

Another crucial aspect of any processor is memory. The human brain much like the processor needs short term memory to function properly; without it humans and computers would be extremely forgetful things. The processor needs short term memory or a place in its brain where it can store data temporarily while its attempting to calculate instruction sets.

The technical term for this type of memory is cache memory. You can identify cache memory on a processor itself as small tiny chunks or blocks soldered directly into the processor. The memory is interconnected through a series of digital circuits and pathways that help the processor get its jobs done by storing data that it's working

L1 Cache Memory

The fastest type of memory in a processor is known as primary cache memory. Also known as L1 cache, this type of memory is the closest to the processor and the fastest. This memory connects to other chunks of memory found in the second level of cache memory labeled L2 cache.

L2 Cache Memory

L2 cache memory acts a temporary buffer zone for the data the flows between the processor, the system memory, and data resident in the L1 cache memory. This is a processor's direct line of communication between system memory and the mainboard's buses. A bus is an electronic pathway the processor can utilize for effective transmission of data to other peripherals.

L2 cache catches data that the L1 layer missed or did not cache and is typically slower than the L1 layer. Very old style mainboards had L2 cache memory equipped on an expansion board (also known as a daughter board that plugged into a special PCI slot connector on the mainboard. Proprietary companies (like Dell, Hewlett Packard, and Gateway) often uses proprietary designs in their computers. Proprietary systems limit the upgrade ability of a computer, and are not my preferred method of choice when building new computers because you are limited to what the hardware manufacturer supports in their designs.

The advent of the Pentium Pro chip allowed for the design of L2 cache memory directly within the processor. At that time, a typical Pentium II processor chip was designed with 512 kilobytes (KB) of L2 cache memory. During this era, Intel also built a line of chips referred to as the Celeron (a less powerful Pentium processor, more suitable for budget based computers). This type of chip came equipped with 128KB of L2 cache, half the amount of a true Pentium processor. The Pentium III processor was later developed and contained the same amount of L2 cache memory as the Pentium II. Processors manufactured by todays standards, are equipped with a minimum amount of 1024 kilobytes of L2 cache. The more L2 cache available the faster the processor can think, and the faster it can finish its duties.

At the time of this writing the standard was 256KB of L2 cache memory for a Celeron processor, 512KB for a Pentium II and III processor, and 1024 KB for a Pentium IV processor.

L3 Cache Memory

For an extra boost in performance modern processors are now equipped with a third layer of cache memory respectively labeled L3 cache. This type of memory resides on a mainboard, and provides more storage between the processor and system memory. L2 and L3 cache memory are essentially the same thing. Newer Pentium IV based processors are equipped with an additional 16 Kilobytes (KB) of memory. Although this may sound like a very tiny amount of memory, it does help speed things along, especially for multimedia intensive applications such as digital video editing.

What's The Time?

Another factor that determines a central processing unit's performance capabilities is its internal clock speed also known as its clock rate. In its most basic sense, clock speed is how fast the processor can actually get things done. In the computer world and statistically speaking, clock rate is a measured in cycles per seconds of how fast the processor can execute instructions.

A Pentium II processor utilizes a clock speed of 100 or 133 Mhz. The Celeron processor in a same class runs at a 66 Mhz clock speed. Pentium III processors run at a 133 Mhz speed, while Pentium IV units run at either a 400 Mhz or 800 Mhz clock speed depending on the class of the unit.

How Is A Processor's Clock Speed Measured?

In prehistoric days when the 80486 ruled technology, the clock speed of a processor  measured in either Megahertz (MHz). At present day low class processors are measured in Gigahertz (Ghz) for medium to high end processors.

For comparison purposes, the first commercial PC released in 1981 came equipped with a processor running at a 2 MHz clock speed (which allowed for 2,000,000 cycles per second to be executed by the processor). A processor chip running at 100 MHz can execute more than 100 million cycles per second). At present day, a standard Pentium IV class processor designed with a 3GHz processor chip, can execute more than three billion cycles per second, a far difference from the first processors engineered.

Bus Speeds

One final characteristic of processors is it's internal front-side bus speed. The phrase front side bus simply refers to the speed at which the processor can electronically transmit data to other peripheral devices inside the computer. The faster the front side bus, the faster the speed at which data can travel.

How Is A Processor's Front Side Bus Speed Measured?

Older technology defined the front side bus speed in Megahertz. A typical Pentium II processor was designed to run at a 100 MHz bus speed while a Pentium III processor ran at 133 MHz, a shade faster than the Pentium II. A budget style Pentium II compatible processor labelled the Celeron, ran at a 66 Mhz front side bus speed. This Celeron chip was a lower style Pentium processor because it contained a smaller amount of cache memory and was best suited for low end workstations.

It is also important to note, that in the processor world, every bit of speed helps, so the difference of 66 MHz equipped in a Celeron processor highly impacts the overall performance of a system and makes it more of a bottlneck for multimedia intensive applications and games. The design of the Celeron slowly changed to allow it to run at a 100 MHz bus speed, but this did not happen until the Celeron was categorized into the Pentium III family of processors. Eventually design specifications permitted the Celeron to run a 266 MHz front side bus, and most Pentium IV classifications of Celeron processors are now equipped to run at either 400 or 533 Mhz bus speed.

A typical Pentium IV processor runs at an internal bus speed of 800 MHz, almost eight times the speed of a Pentium III chip.

AMD was not far behind in the design of their processors. They released equally powerful processors to keep current in the marketplace. They first released two processors, one labelled the K6-II and another labeled the K6-III. Both chips ran at a 100 MHz front side bus speed. To keep up with the technologies, AMD later released a Pentium III comparable chip entitled the Athlon K7 which performed at a 200 MHz bus speed. Shortly thereafter the processor manufacturer released a Pentium IV comparable chip entitled the G5 which ran between a 900 MHz and 1.25 GHz bus speed.

Summary

You should understand by now that there are several different factors that affect the performance and architecture of a particular brand of chip including the amount of cache memory included on the chip, the speed at which it can calculate data, as well as the speed at which it can communicate with other devices in the computer. The faster the clock speed does not necessarily mean a better chip, it just means that it will perform a shade faster than its competitors. The real performance impact of any processor combines the quantity of cache memory on the chip, and the clock speed at which it can run.