Processors
Processors, or CPU’s, are the central processing units of the computer. They perform all of the calculations necessary for normal operation. They also control a large number of I/O functions. Locating the CPU is rather easy – look for a fan on the motherboard rising high above the motherboard and underneath that fan and heatsink will lay a large chip. That chip is the CPU of the computer. Most PC’s utilize a single CPU. However, a PC may utilize more than one CPU, usually meaning it is a dual CPU computer. Even still, there are some machines that can manage four processors. For the purposes of this guide, we will focus on individual processors, as server-oriented dual processors such as Intel’s Xeon and Itanium and AMD’s Athlon MP are not focused upon in this test.
There are several ways to classify processors. The first, major way is by the family of processor. There exist several major families of processors, including the 80X86, Pentiums I, II, III, and IV, and various lines of AMD processors. You will find that the A+ test focuses very much on typical IBM architecture, which up until recently was built entirely around the Intel CPU’s. Therefore, most questions dealing with CPU’s and such on this test are focused upon the Intel architecture. However, even within the family of chips, there are a number of discerning characteristics between processors. Some processors are derivatives of other processors, and add on or subtract features from the original in some kind of marketing scheme. In addition, these derivative chips differ in clock speed with one another. A major testing point seems to also be the presence or lack thereof of a floating point unit, or FPU. Another major testing point concerns the presence and amount of internal cache. Finally, it is important to determine the width of the data paths in the CPU, both internal and external, as described previously.
80X86 Processors
The 80X86 processors are considered the first modern CPU’s for IBM computers. Designed and manufactured by Intel, the 80X86 line began with the 8088 and ended with the 80486 derivative processors. Each of these added on features, quality, and clock speed.
The 8088 Processor
The Intel 8088 processor chip was technically not the first IBM compatible CPU in existence. Rather, the 8086 was the first developed IBM CPU for the mass market. However, for some marketing reason, the 8088 was the first to hit the market. It is thus the traditional starting point in the history of IBM CPU’s. It was developed and designed for the XT architecture, and clocked at about 4.77 MHz. For those of you who are not familiar with the term hertz, one hertz is a tick, or cycle, in a second. As the number of hertz increases, a proportionally larger associated number of ticks per second results, thus associating hertz with speed. One Megahertz, or MHz, is 1024 squared number of ticks per second. This is roughly equal to a million ticks per second (the actual number is a bit greater – 1,048,576). Later, it was clocked at 8 MHz, roughly double of the original speed. It featured a 16-bit internal and 8-bit external bus, which fit nicely with XT model architecture. It also had a 20-bit address bus to access 1 MB of RAM. To finish execute an instruction, the 8088 took about 12 cycles, on average.
The Failed 8086
Recall that the first processor was the 8086. However, due to an issue in pricing, Intel and IBM felt that the majority of consumers would not pay the extra money to achieve the features associated with the 8086, including a true 16-bit internal bus and a true 16-bit external bus. As it was able to communicate at 16 bits with other components of the computer, it was able to achieve a 20% performance gain over the 8088.
Partial Success: The 80186 and the 80188
Both the 80186 and the 80188 suffered similar problems. Both still addressed memory at 20-bits, limiting the potential RAM to be used to only a single megabyte. Additionally, although they combined much system I/O components, most consumers and PC manufacturers did not believe these gains constituted a purchase. However, some consumers appreciated the integration and purchased these processors.
The 80286 Processor
The 80286 processor represents a turning point in computing. First, the release of the AT motherboard coincided with the release of this processor (IBM and Intel cooperation) and thus this combination became to be referred to as the IBM clone. This association lead to the standardization of the AT motherboard. Moreover, the 80286 was able to address 16 entire megabytes of memory – nothing by today’s standards, but a revolutionary achievement at the time. However, the way programs are written are highly specific in the means they access memory, and thus the processor instruction codes for these programs differ when there is a different mechanism for addressing memory. Intel wisely made the 80286 capable of running older 8088/8086 compatible programs which were designed to only access a single megabyte of memory. In addition, the 80286 required on average 4.5 cycles per instruction, rather than the 12 required by the 8088/8086 processors. The 80286 was also available in many clock speeds, including 8, 10, 12, and 20 MHz. Even so, the 80286 would install on motherboards limited at 16 MHz, even when clocked at a higher speed.
The most remarkable achievement of the 80286, however, was the creation of real mode. The 80286, which is capable of addressing 16 megabytes of memory, was the first chip to introduce both “Real” and “Protected” mode. Real Mode is the mode in which the processor addresses the first 1,024 bytes of memory as conventional and thus assigns them real locations, allowing backwards compatibility. The 286 was capable of addressing 16 megabytes of memory, as previously mentioned, though it could extend that number through a feature known as swapping. This feature would prove to be indeed important in the future of computing. Swapping memory is the process of using the hard drive as a holding space for RAM memory. In other words, if the system requires more RAM, the system can utilize the hard drive to copy blocks of memory to the hard drive, use those blocks, and then eventually, when those blocks are needed, return them to the physical memory addresses. Through swapping, software could access up to one gigabyte, or 1024 cubed, bytes of memory. However, swapping was only available in “Protected Mode,” which was not fully utilized until later. Protected Mode, theoretically, allows a program to crash, or stop responding tot the CPU, without causing the entire system to fail. This is because in Protected Mode, each of the memory addresses was fully independent of one another. This also allowed the CPU to swap memory to the hard drive. However, in practice, software developers would not make use of swapping till a bit later, and even today, under a Pentium IV and Windows XP, one program can still cause the entire system to crash. Despite the realities, Protected Mode was a very important development in the CPU.
The PCMech.com weekly newsletter has been running strong for over 8 years. Sign up to get tech news, updates and exclusive content - right in your inbox. Also get (several) free gifts.
