On October 19, 1992, Intel announced that the fifth generation of its compatible microprocessor line (codenamed P5) would be named the Pentium processor rather than the 586, as everybody had assumed.

Calling the new chip the 586 would have been natural, but Intel discovered that it could not trademark a number designation, and the company wanted to prevent other manufacturers from using the same name for any clone chips they might develop. The actual Pentium chip shipped on March 22, 1993.

Systems that used these chips were only a few months behind. The Pentium is fully compatible with previous Intel processors, but it differs from them in many ways. At least one of these differences is revolutionary: The Pentium features twin data pipelines, which enable it to execute two instructions at the same time.

The 486 and all preceding chips can perform only a single instruction at a time. Intel calls the capability to execute two instructions at the same time superscalar technology. This technology provides additional performance compared with the 486.

With superscalar technology, the Pentium can execute many instructions at a rate of two instructions per cycle. Superscalar architecture usually is associated with high-output RISC chips. The Pentium is one of the first CISC chips to be considered superscalar. The Pentium is almost like having two 486 chips under the hood.

The two instruction pipelines within the chip are called the u- and v-pipes. The u-pipe, which is the primary pipe, can execute all integer and floating-point instructions. The v-pipe is a secondary pipe that can execute only simple integer instructions and certain floating-point instructions.

The process of operating on two instructions simultaneously in the different pipes is called pairing. Not all sequentially executing instructions can be paired, and when pairing is not possible, only the u-pipe is used. To optimize the Pentium's efficiency, you can recompile software to enable more instructions to be paired.

The Pentium processor has a branch target buffer (BTB), which employs a technique called branch prediction. It minimizes stalls in one or more of the pipes caused by delays in fetching instructions that branch to nonlinear memory locations.

The BTB attempts to predict whether a program branch will be taken and then fetches the appropriate instructions. The use of branch prediction enables the Pentium to keep both pipelines operating at full speed. The Pentium has a 32-bit address bus width, giving it the same 4GB memory-addressing capabilities as the 386DX and 486 processors.

But the Pentium expands the data bus to 64 bits, which means it can move twice as much data into or out of the CPU, compared with a 486 of the same clock speed. The 64-bit data bus requires that system memory be accessed 64 bits wide, so each bank of memory is 64 bits.

On most motherboards, memory is installed via SIMMs or DIMMs. SIMMs are available in 8-bit-wide and 32-bit-wide versions, whereas DIMMs are 64 bits wide. In addition, versions are available with additional bits for parity or error correcting code (ECC) data.

Most Pentium systems use the 32-bit-wide SIMMs—two of these SIMMs per bank of memory. Most Pentium motherboards have at least four of these 32-bit SIMM sockets, providing for a total of two banks of memory.

Later Pentium systems and most Pentium II systems still in use today use DIMMs, which are 64 bits wide—just like the processor's external data bus, so only one DIMM is used per bank. This makes installing or upgrading memory much easier because DIMMs can go in one at a time and don't have to be matched up in pairs.

Even though the Pentium has a 64-bit data bus that transfers information 64 bits at a time into and out of the processor, the Pentium has only 32-bit internal registers. As instructions are being processed internally, they are broken down into 32-bit instructions and data elements and processed in much the same way as in the 486.

Some people thought that Intel was misleading them by calling the Pentium a 64-bit processor, but 64-bit transfers do indeed take place. Internally, however, the Pentium has 32-bit registers that are fully compatible with the 486. The Pentium has two separate internal 8KB caches, compared with a single 8KB or 16KB cache in the 486.

The cache-controller circuitry and the cache memory are embedded in the CPU chip. The cache mirrors the information in normal RAM by keeping a copy of the data and code from different memory locations. The Pentium cache also can hold information to be written to memory when the load on the CPU and other system components is less. (The 486 makes all memory writes immediately.)

The separate code and data caches are organized in a two-way set associative fashion, with each set split into lines of 32 bytes each. Each cache has a dedicated translation lookaside buffer (TLB) that translates linear addresses to physical addresses. You can configure the data cache as write-back or write-through on a line-by-line basis.

When you use the write-back capability, the cache can store write operations and reads, further improving performance over read-only write-through mode. Using write-back mode results in less activity between the CPU and system memory—an important improvement because CPU access to system memory is a bottleneck on fast systems.

The code cache is an inherently write-protected cache because it contains only execution instructions and not data, which is updated. Because burst cycles are used, the cache data can be read or written very quickly. Systems based on the Pentium can benefit greatly from secondary processor caches (L2), which usually consist of up to 512KB or more of extremely fast (15ns or less) SRAM chips.

When the CPU fetches data that is not already available in its internal processor (L1) cache, wait states slow the CPU. If the data already is in the secondary processor cache, however, the CPU can go ahead with its work without pausing for wait states.

The Pentium uses a Bipolar Complementary Metal-Oxide Semiconductor (BiCMOS) process and superscalar architecture to achieve the high level of performance expected from the chip. BiCMOS adds about 10% to the complexity of the chip design, but adds about 30%–35% better performance without a size or power penalty.

All Pentium processors are SL enhanced—they incorporate the SMM to provide full control of power-management features, which helps reduce power consumption. The second-generation Pentium processors (75MHz and faster) incorporate a more advanced form of SMM that includes processor clock control.

This enables you to throttle the processor up or down to control power use. You can even stop the clock with these more advanced Pentium processors, putting the processor in a state of suspension that requires very little power. The second-generation Pentium processors run on 3.3V power (instead of 5V), reducing power requirements and heat generation even further.

Many Pentium motherboards supply either 3.465V or 3.3V. The 3.465V setting is called VRE (voltage reduced extended) by Intel and is required by some versions of the Pentium, particularly some of the 100MHz versions. The standard 3.3V setting is called STD (standard), which most of the second-generation Pentiums use.

STD voltage means anything in a range from 3.135V to 3.465V with 3.3V nominal. Additionally, a special 3.3V setting called VR (voltage reduced) reduces the range from 3.300V to 3.465V with 3.38V nominal. Some of the processors require this narrower specification, which most motherboards provide.

For even lower power consumption, Intel introduced special Pentium processors with voltage reduction technology in the 75 to 266MHz family; the processors were intended for mobile computer applications. They did not use a conventional chip package and were instead mounted using a new format called tape carrier packaging (TCP).

The tape carrier packaging does not encase the chip in ceramic or plastic as with a conventional chip package, but instead covers the actual processor die directly with a thin, protective plastic coating. The entire processor is less than 1mm thick, or about half the thickness of a dime, and weighs less than 1 gram.

They were sold to system manufacturers in a roll that looks very much like a filmstrip. The TCP processor is directly affixed (soldered) to the motherboard by a special machine, resulting in a smaller package, lower height, better thermal transfer, and lower power consumption.

Special solder plugs on the circuit board located directly under the processor draw heat away and provide better cooling in the tight confines of a typical notebook or laptop system—no cooling fans are required. For more information on mobile processors and systems, see the chapter "Portable PCs" included on the DVD with this book.

The Pentium, like the 486, contains an internal math coprocessor or FPU. The FPU in the Pentium was rewritten to perform significantly better than the FPU in the 486 yet still be fully compatible with the 486 and 387 math coprocessors. The Pentium FPU is estimated to be two to as much as ten times faster than the FPU in the 486.

In addition, the two standard instruction pipelines in the Pentium provide two units to handle standard integer math. (The math coprocessor handles only more complex calculations.) Other processors, such as the 486, have only a single-standard execution pipe and one integer math unit. Interestingly, the Pentium FPU contains a flaw that received widespread publicity.