The Intel Processor Lineup

Over time, Intel has introduced several generations of microprocessors. Each processor type is referred to as a generation; each is based on the new technological enhancements of the day. With each product release come new software and hardware products to take advantage of the new technology.

Several generations of Intel processors are available today. Since the arrival of the first Intel chip in the IBM PC, Intel has dominated the market. It seems that every time you turn around, a new chip promises greater performance and processing capabilities than the previous one.

What makes Intel the market leader is its ability to bring the newest innovations in chip technology to the public, usually before its competitors, who are not far behind. Competition is fierce, and each manufacturer attempts to improve on the designs of the others, releasing similar chips that promise better performance.

The following table shows the specifications for the Intel processors issued to date. You should read the specifications and reviews of each processor to understand its capabilities and reliability.

Model	Year Introduced	Maximum Internal Clock Frequency	Data Path(in Bits)
4004	1971	.108MHz	4
8008	1972	.2MHz	8
8080	1974	2MHz	8
8086	1978	8MHz	16
8088	1979	4.77MHz	8
80286	1982	20MHz	16
80386	1985	40MHz	32
80486	1989	100MHz	32
Pentium	1993	200MHz	32
Pentium Pro	1995	200MHz	32
Pentium MMX	1997	233MHz	32
Pentium II	1997	450MHz	64
Pentium II Xeon	1998	450MHz	64
Celeron	1998	500MHz	64
Pentium III	1999	733MHz	64
Pentium III Xeon	1999	550MHz	64
Itanium	2000	800MHz	64
Tualatin	2001	1.2GHz	64
Pentium 4	2000	2.2GHz	64

Factors Affecting Performance

Many factors come together to determine the performance of any computer. All other factors being equal, faster components will give better performance, but any computer will be limited by its 'weakest links.' As an analogy, consider that putting a larger engine in a standard automobile will make it faster, but only if the automobile is going in a straight line. As soon as you try to make the car follow a twisting road, other components such as the drivetrain and the tires can limit the performance of the larger engine.

Within a processor family, faster processors will outperform slower processors. But when we're comparing processors from different families, that rule does not apply. For example, the rating of 400MHz for a processor from one family does not indicate that it will run significantly faster than a 333MHz processor from a more advanced processor family.

As you learned earlier, clock cycles and data path are two factors that can influence the performance of your computer. Other factors are:

Cache memory  Very fast memory that sits between the CPU and the main RAM. Cache memory can be as fast as 5 to 10 nanoseconds, whereas main RAM is usually not faster than 60 to 70 nanoseconds. (Yes, a lower number is better here because it indicates that the memory takes less time to move data.)

Bus speed  The rate at which data can be transferred between the CPU and the rest of the motherboard. Typical bus speeds are 733MHz and higher with the current standard for motherboards entering the market.

Note 

The type of peripherals on your computer can affect system performance. If your application spends a lot of time accessing your hard disk, selecting a better-performing disk system would improve CPU efficiency. For example, Small Computer System Interface (SCSI) hard disks place a much smaller overhead burden on your CPU than Integrated Device Electronics (IDE) storage devices. Storage systems are covered in detail in Chapter 2, 'Storing Your Files: Data Storage.'

History of Intel Chips

Intel released the world's first microprocessor, the Intel 4004, in 1971. It was a 4-bit microprocessor containing a programmable controller chip that could process 45 instructions. The 4 bits meant that the chip had four lines for data to travel on, much like a four-lane highway. Because of its limitations, it was implemented only in a few early video games and some other devices. The following year, Intel released the 8008, an 8-bit microprocessor with enhanced memory storage and the capability to process 48 instructions.

Intel then began to research and develop faster, more capable processors. From that research emerged the 8080, which could process instructions 10 times faster than its predecessors. Although the speed had dramatically improved, it was still limited by the number of instructions it could process. Finally, in 1978, Intel broke many barriers by releasing the first of many computer-ready microprocessors, the 8086. The 8086 was a breakthrough technology with a bus speed of 16 bits and the capability to support and use 1MB of RAM. Unfortunately, the cost of manufacturing such a chip and compatible 16-bit components made the chip unaffordable. Intel responded the following year with the production of an 8-bit chip, the 8088.

millions of instructions per second (MIPS)

A measurement of the number of microcode instructions that a CPU or microprocessor can complete in one second, or cycle.

protected mode

A mode available in Intel 80286 and 80386 processors. Added the capability for the processor to allocate each application its own separate memory space. In the event that an application crashed, the rest of the system was protected.

multitask

To perform several operations concurrently.

Intel continued to break new ground as the release of each new generation of processor offered improved functions and processing capabilities. The most dramatic improvement was the number of instructions, based on a scale of millions, that the processor could process in one second. This rate, referred to as millions of instructions per second (MIPS), ranges from 0.75MIPS for the 8088 to over 450MIPS for Pentium 4 processors.

The second most dramatic improvement was the speed of the internal clock, measured in megahertz. All processors are driven by an internal clock mechanism that keeps the rhythm of the chip, much like the rhythm of a heartbeat. The faster the speed of the internal clock, the faster the processor can process instructions. Intel continued to increase the speed of the internal clock from 4.77MHz for the original 8088 to more than 2.2GHz for the newest generation of Intel microprocessors.

The First Generation: 8086 and 8088

The first major processor release from Intel was the 8086 microprocessor. The processor debuted as the first evolutionary step in a multitude of processors, each improving on the design of the original 8086. The lineage is referred to as the x86 family of microprocessors. Although this first release was crude compared to today's standards, it paved the way for the others to follow.

The 8088 was released a short time after the 8086 but was not as powerful as its predecessor. The 8086, a true 16-bit processor, contained 16-bit registers and a 16-bit data path. Motherboard technology had not quite reached the 16-bit level and was still costly in 1981. IBM decided to use a version of the 8088 chip, with the same design but with an 8-bit data path to accommodate the widely used 8-bit technology of the time.

The Second Generation: 80286

Intel forged a new milestone in PC processor technology with the release of the 80286, more commonly called the 286. The 286 offered a significant performance increase over the 8086 and 8088 with the unique capability to operate in a protected mode. The protected mode enabled the processor to multitask and still included its normal, or real, mode of operation.

Real mode required memory to be accessed in a linear format. This means that data being sent to RAM had to be placed in the order it was received-one application after another. With this limitation, instructions were usually processed one at a time.

transistor

A microscopic electronic device that uses positive electrons to create the binary value of one, or 'on,' and negative electrons to create the binary value of zero, or 'off.' Modern CPUs have millions of transistors.

conventional memory

The first 640KB of memory, which is required by DOS to run. Memory above 640KB was used by operating systems such as Windows3.1.

DOS

An operating system developed by Microsoft. DOS predominantly uses command lines to manage the operating system, applications, and files.

Windows operating system

An operating system developed by Microsoft that provides a graphical user interface (GUI) for DOS.

Protected mode enabled multitasking to occur by allocating a specific range within memory for each task. Applications could therefore be accessed simultaneously, greatly improving performance.

Note 

Some companies produce upgrades for 286, 386, and 486 computers. The processor upgrades are relatively inexpensive and can greatly improve the overall CPU speed. Although 286 upgrades are nearly impossible to find, upgrades for 486-based computers are available. The processor upgrade can convert your aging 486 into a speedy computer, comparable to a Pentium.

The Third Generation: 80386

Intel's introduction of the 80386 processor reached yet a new milestone, condensing more than 250,000 transistors onto a single 32-bit processor chip. (The number of transistors on a processor is an indicator of the complexity of the processor and of its ability to perform complex calculations.) This new generation of processors incorporated true, fully functioning multitasking capabilities. Protected mode was now commonly referred to as the 386 Enhanced Mode, because the 80386 was able to overcome the multitasking limitations of the 80286.

The 80386 included a new operating mode called Virtual Real Mode. This new mode created conventional memory space required by DOS programs to run within the Windows operating system. Virtual Real Mode, or Virtual DOS Mode as it is commonly called, is still used for running DOS-based games and applications within Windows 95 and 98.

Note 

You will learn more about DOS and Windows operating systems in Chapter 5, 'Desktop Operating Systems: A Comparison,' and Chapter 6, 'DOS 101: DOS Basics Every MCSA and MCSE Should Know.'

Several types of 80386 chips were issued, each with a unique combination of features. Intel offered two options: the 80386DX and the 80386SX CPUs. Both were 32-bit processors, but the 80386DX used a 32-bit data path and the 80386SX used a 16-bit data path. Although the SX chip had smaller data paths, it was more competitively priced.

Peripheral Component Interconnect (PCI)

A bus standard for the transfer of data between the CPU, expansion cards, and RAM. PCI communicates at 33MHz.

bus architecture

Any linear pathway on which electrical signals travel and carry data from a source to a destination.

pipeline

A place in the processor where operations occur in a series of stages. The operation is not complete until it has passed through all stages.

The Fourth Generation: 80486

Like the 80386, the next family of processors was released in 80486SX and 80486DX versions. Both included a 32-bit internal and external data path and an original internal clock frequency of 33MHz. The SX version was released with the numeric processor, or FPU, disabled and the internal clock speed slowed to 20MHz to offer a lower-cost processor to the consumer. Later this became a limitation with the emergence of more powerful software applications. A numeric processor was issued to complement the SX, turning it into a fully functioning DX.

A dramatic improvement was engineered into later deployments of the processor. A mechanism called a clock doubler enabled the internal system clock to run at twice the normal bus speed. Soon the 486DX-33 became the 486DX2-66, with the 2 signifying the clock-doubling technology. Eventually the idea of increasing the clock speed led to a clock tripler.

The Next Generations: The Pentium Family

Intel released the Pentium chip to take advantage of the newly released Peripheral Component Interconnect (PCI) bus architecture. The new processor also consisted of 3.1 million transistors and a new 64-bit data path. The chip was originally designed to operate at 66MHz but was scaled down to 60MHz to support the new transistor design, which was experiencing heat and power problems. The first chips deployed also suffered from a bug in the microcode that hampered the processor's capability to calculate complex mathematical equations with precision. This problem was immediately fixed through a new batch of chips.

The most significant development in the Pentium was the use of two parallel 32-bit pipelines that enabled it to execute twice the number of instructions as previous Intel processors-a technological advancement that Intel named superscalar technology. Almost all processors today use this technology.

Released with the Pentium family of processors was Multimedia Extension (MMX) technology. MMX technology is often referred to as multimedia-enhanced technology, but this is not completely accurate. MMX-equipped processors contained additional instruction code sets that increased the processing speed for audio, video, and graphical data by up to 60 percent as compared to traditional Pentium processors. The MMX chips dramatically improved the response time of games and multimedia-based applications.

Multimedia Extension (MMX)

A processor technology that dramatically improves the response time of games and multimedia-based applications. The technology was introduced through the MMX-equipped line of Intel Pentium chips.

The types of Pentium processors include:

• Pentium

• Pentium MMX

• Pentium Pro

• Pentium II (PII)

• Celeron

• Pentium II Xeon

• Pentium III (PIII)

• Pentium III Xeon

• Itanium/Itanium 2

• Pentium 4

• Tualatin

Pentium

The Pentium chip introduced the world to the first parallel 32-bit data path, which enabled the Pentium to process 64 bits-twice as much data as before. The Pentium was the first microprocessor chip designed to work with the PCI bus specification and had internal clock speeds ranging from 60MHz to 200MHz.

Pentium MMX

The Pentium with MMX technology included an expanded instruction code set with 57 new MMX microcode instructions. MMX enabled the microprocessor to increase the processing speed of audio, video, and graphics by up to 60 percent.

Pentium Pro

The Pentium Pro was the successor to Intel's Pentium processor. One of the unique features of this microprocessor was its internal RISC architecture with CISC-RISC translator service. The translator service was able to use the CISC set of instructions, common to all Intel chips, convert them to the RISC set, the faster of the two, and then complete the tasks as necessary using RISC.

The architectural enhancement that really distinguished the Pro from the original Pentium would influence how most microprocessors would later be developed. The Pro was two chips in one: On the bottom of the Pentium was the actual processor. Connected directly overhead of the processor was an L2 cache. By placing the L2 cache close to the processor, Intel was able to greatly increase the performance of the Pentium Pro.

Pentium II (PII)

The Intel Pentium II, or PII, processor is essentially an enhanced Pentium Pro processor with MMX extensions, cache memory, and a new interface design. The PII was designed to fit into an SEC that plugs into a 242-pin slot.

Celeron

The only noticeable difference between the Celeron and regular Pentium II processors is the lack of cache memory within its cartridge. Later models of the Celeron include cache memory on the same chip as the processor.

Pentium II Xeon

One of the major enhancements in the Pentium II Xeon is larger on-board cache. This processor is available with either 1MB or 2MB of L2 cache and a clock rate of 450MHz.

Pentium III (PIII)

With its faster clock rates (up to 733MHz), the Pentium III supports demanding applications such as full-screen, full-motion video and realistic graphics. Seventy new instructions have also been added to make technologies such as 3-D graphics, video, speech, and imaging faster and more affordable for mainstream users. Each Pentium III also contains a unique processor serial number. Intel's intent behind this feature was to enhance system security and asset tracking. However, many individuals object to the serial number as infringement on their privacy because it can be used to identify computers on the Internet.

Pentium III Xeon

The Pentium III Xeon processor challenges RISC-based servers in both price/performance and raw performance. It is available in speeds of up to 550MHz and supports configurations that have more than one processor in the same box.

Itanium/Itanium 2

Previously known by the code name Merced, the Itanium processor employs a 64-bit architecture and enhanced instruction handling to greatly increase the performance of computational and multimedia operations, and supports clock speeds of up to 800MHz. The Itanium 2 processor uses a 128-bit architecture and supports speeds of 900MHz and 1GHz.

Pentium 4

The latest family of Intel processors is the Pentium 4, which have been marketed to the advanced desktop market. The architectural changes allow the processor to increase performance by processing more instructions per clock cycle. This technology is referred to as Hyper Pipeline and allows for 20 pipeline stages as opposed to 10 pipeline stages used in the Pentium III family. Other enhancements are added through NetBurst technology, which includes such features as improved L1 and L2 caches and the Rapid Execution Engine. Current Pentium 4 processors can support speeds of up to 2.53GHz.

Tualatin

The Tualatin processor was originally designed to be a logical progression to the Pentium III family. However, as the schedule for this processor slipped, Intel shifted focus to the Pentium 4 processor family. As a result, the Tualatin processor was not released until mid-2001. The Tualatin processors support speeds of up to 1.2GHz.

Warning 

Even though the Tualatin processor appears to be compliant with Socket 370, the clocking, voltage, and signal levels it uses make it incompatible with existing Pentium III motherboards.

Note 

Although most of us would like to get our hands on the new high-speed processors, the reality is that it will be a while before they are affordable. Also, to really reap the benefits of those high-speed CPUs, computers will need to have equivalently high-powered hardware. That is why you will see the first high-speed high CPUs only in expensive servers.