Wintel

Wintel is a colloquial, often pejorative, term used to describe industry standard desktop computers of the type commonly used in homes and businesses in the wealthier parts of the world since the late 1980s, which usually combine the Microsoft Windows operating environment system with Intel's x86 processor architecture.

The term is a concatenation of Windows (Microsoft's Windows operating environment) and Intel (the largest manufacturer of CPUs and the originator of the X86 processor architecture. In a strict sense, Wintel refers to only the two-thirds of desktop computers that run Windows on an Intel processor, but it is often generalised so that, for example, a machine running Linux on an AMD CPU is, in this broader sense, still "Wintel". In this broader meaning, around 95% of all personal computers currently in service are "Wintel boxes".

Background

Even more so than the mini-computer[?] market which preceeded it, the micro-computer[?] market when it began in the late 1970s was led by a host of small, innovative companies, often start-ups. With the market proved by the earliest manufacturers, larger compaies soon began to take a hand, sometimes successfully (Tandy, Hewlett-Packard), sometimes not (Texas Instruments). One by one, the majors in the "big iron" market—mainframe and mini-computer makers—recognised the emergent boom in the micro-computer market.

By the early 1980s, the chaos and incompatibility of the first years had given way to a smaller number of industry standards, including the S-100 bus, CP/M, the Apple II, Microsoft BASIC in ROM, and the 5.25 inch floppy drive. Despite the presence of informal standards which allowed a fair measure of interoperability between different machines from different manufacturers, no single company controlled the industry, and fierce competition ensured that innovation in both hardware and software was the rule rather than the exception. Most of the software used today is directly derived from the ideas that grew out of this creative bonanza—the obvious example is the spreadsheet, but there are countless others.

In 1981 the largest and oldest computer firm of them all, IBM, finally entered the microcomputer market, with a machine that was created by a small sub-division of the firm and was very unusual by their standards, insofar as it was largely sourced from outside component suppliers, technically unambitious, very similar to a number of existing 16-bit 8088-based micro-computers, ran third-party operating systems, and above all, had an open architecture[?]. It was called the IBM PC and it became the most successful single computer of all time.

The key feature of the IBM PC was that despite its technical mediocrity and higher than market price, it had IBM's enormous public respect behind it. It was an accident of history that the IBM PC happened to have an Intel CPU (instead of the technically superior Motorola 68000 that had been tipped for it, or an IBM in-house design), and that it shipped with IBM PC-DOS (a licenced version of Microsoft's MS-DOS rather than the technically superior industry-standard CP/M-86 operating system), but an accident that was to have enormous significance in later years.

Because the IBM PC was an IBM product with the IBM badge, personal computers (as they were beginning to be called) became respectable. It became much easier for a conservative business to justify buying a microcomputer than it had been even a year or two before—and easiest of all to justify buying the IBM personal computer.

Because the architecture was open (an idea pioneered by Apple in the early years but suddenly abandoned by them when the Apple II gave way to the Mac) it was possible for third-party manufacturers to design and market add-in cards to bring new functions or enhance existing ones. Within a couple of years there were literally thousands of different third-party add-in cards available for every imaginable purpose, from hundreds of manufacturers both large and small.

Industry competitors took one of several approaches to the changing market. Some (such as Apple, Atari, and Acorn) perservered with their independent and quite different systems. Others (notably world number two Digital, Hewlett-Packard, and Apricot[?]) concentrated on making similar but technically superior models. Other early market leaders (such as Tandy-Radio Shack, Texas Instruments and Commodore) stayed with outdated architectures and proprietary operating systems for some time before belatedly realising which way the market wind was blowing and switching to the most successful long-term business strategy, which was to build a machine that duplicated the IBM PC as closely as possible and sell it for a slightly lower price, or with higher performance. Given the very conservative engineering of the early IBM personal computers and their higher than average prices, this was not a terribly difficult task at first, bar only the great technical challenge of crafting a BIOS that duplicated the function of the IBM BIOS exactly but did not infringe on copyrights.

The two early leaders in this last strategy were both start-up companies: Columbia Computers[?] and Compaq. They were the first to achieve reputations for very close compatibility with the IBM machines, which meant that they could run software written for the IBM machine without recompilation. Before long, IBM had the best-selling personal computer in the world and at least two of the next-best sellers were, for practical purposes, identical.

For the software industry, the effect was profound. First, it meant that it was rational to write for the IBM PC and its clones as a high priority, and port versions for less common systems at leisure. Second (and even more importantly), where software written in pre-IBM days had to be careful to use as plain a sub-set of the possible techniques as practicable (so as to be able to run on any hardware that ran CP/M), with a major part of the market now all using the same exact hardware (or a very similar clone of it) it was practical to take advantage of any and every hardware-specific feature offered by the IBM. Independent BIOS companies like Award, Chips & Technologies, and Phoenix began to market clean room BIOS that was 100% compatible with IBM's, and from that time on any competent computer manufacturer could achieve IBM compatibility as a matter of routine.

From around 1984, the market was fast growing but relatively stable. There was as yet no sign of the "Win" half of "Wintel", though Microsoft were achieving enormous revenues from DOS sales both to IBM and to an ever-growing list of other manufacturers who were illegally forced to buy an MS-DOS licence for every machine they made, even those that shipped with competing products! As for Intel, every PC made either had an Intel processor or one made by a second source supplier under licence from Intel. Intel and Microsoft had enormous revenues, Compaq and a thousand other makers between them made far more machines than IBM, but the power to decide the shape of the personal computer rested firmly in IBM's hards.

The rise of Wintel

In 1987, IBM made a bold and ultimately disasterous business decision. Although the open architecture of the PC and its succesors had been a great success for them, and they were the biggest single manufacturer, most of the market was buying faster and cheaper IBM compatible machines made by other firms. IBM chose to introduce their PS/2[?] line. The PS/2s remained software compatible, but the hardware was quite different. It introduced the technically superior Microchannel bus for higher speed communication within the system, but failed to maintain the open AT bus (later called the ISA bus), which meant that none of the millions of existing add-in cards would function. The new IBM machines, in other words, were not IBM compatible!

In addition, IBM planned the PS/2 in such a way that for both technical and legal reasons it would be very difficult to clone. Instead, IBM offered to sell a PS/2 licence to anyone who could afford the royalty—but that they would not only require a royalty for every PS/2 compatible machine sold, but also a payment for every IBM compatible machine the particular maker had ever made in the past.

Many PC manufacturers signed up as PS/2 licencees. (Apricot, who had lost badly by perserving with their "better PC than IBM" strategy up until this time, was one of them, but there were many others.) Many others decided to hold off before commiting themselves. Some major manufacturers, known as the Gang of Seven decided to group together and decide on a bus type that would be open to all manufacturers, as fast or faster than IBM's Microchannel, and yet still retain backward compatibility with ISA.

This was the crucial turning point: the industry as a whole was no longer content to let IBM make all the major decisions about technical direction. In the event, the new EISA bus was itself a commercial failure: by the time it was finalised the much cheaper VESA local bus[?] had removed most of the need for it, the Lotus Intel Microsoft[?] joint memory standard had taken RAM compatibility out of the equation, and Intel's PCI bus was just around the corner. But although very few EISA systems were sold, it had achieved its purpose: IBM no longer controlled the computer industry.

At around this same time, the end of the 1980s and the beginning of the 1990s, Microsoft's Windows operating environment started to become popular, and Microsoft's competitor Digital Research started to recover a share of the DOS market. (See DR-DOS.) IBM planned to replace DOS with the vastly superior OS/2 (originally an IBM/Microsft joint venture, and unlike the PS/2 hardware, highly backward compatible), but Microsoft preffered to push the industry in the direction of its own product, Windows. With IBM suffering its greatest ever public humiliation in the wake of the PS/2 disaster, massive financial losses, and a marked lack of company unity or direction, Microsoft's combination of a soft marketing voice and a big financial stick was effective: Windows became the de-facto standard.

For the the competing computer manufacturers, large or small, the only common factors to provide joint technical leadership were operating software from Microsoft, and CPUs from Intel.

The Wintel standard

Over the following years, both firms in the Wintel partnership would attempt to extend their monopolies. Intel made a successful major push into the motherboard and chipset markets (becoming the largest motherboard manufacturer in the world and, at one stage, almost the only chipset manufacturer), but badly fumbled its attempt to move into the graphics chip market, and (from 1991) faced sharp competition in its core CPU territory from AMD and Cyrix.

Microsoft fared better. In 1990, Microsoft had two competitors in its core market (Digital Research and IBM), Intel had none. By 1996, Intel had two competitors in its core market (CPUs), while Microsoft had none. The integration of DOS into Windows 95 was the masterstroke: not only were the other operating system vendors frozen out, Micrsoft could now require computer manufacturers to comply with its demands on pain of higher prices (as when it required IBM to stop actively marketing OS/2 or else pay more than twice as much for Windows 95 as its competitor Compaq) or by witholding "approved for Windows 95" endorsement (which was regarded as an essential hardware marketing tool). Microsoft were also able to require that free publicity be given over to them by hardware makers. (For example, the advertising symbols on all modern keyboards, or the strict licence restrictions on what may or may not be displayed during system boot and on the Windows desktop.) And, of course, Microsoft were able to take over most of the networking market (formerly the domain of Lantastic and Novel) with Windows NT, and the business application market (formerly led by Lotus and WordPerfect) with Office.

Although Microsoft is by far the dominant player in the Wintel partnership now, Intel's continuing influence should not be underestimated. Intel and Microsoft, once the closest of partners, have operated at an uneasy distance from one another since their first major dispute (which was to do with Intel's heavy investment in the 32-bit optimised Pentium Pro and Microsoft's delivery of an unexpectedly high proportion of 16-bit code in Windows 95). Both firms flirt with one-another's competitors from time to time, most notably with Microsoft's close relationship with AMD.

to do: write stuff about Microsoft/Intel cooperation, and recent tensions between them

The Wintel platform is the dominant desktop and laptop computer architecture, and has now spread to be a leading architecture in the server space.

Alternatives to Wintel

Although there are many alternatives to Wintel, most have only tiny market share. The two leading alternatives to Wintel on the desktop and small server are

• Linux and other free Unix-like operating systems such as FreeBSD and OpenBSD, on a variety of architectures, but in particular commodity x86 computers
• Apple computer's Macintosh computers with PowerPC processors

Other markets such as palmtops[?] and mobile phones have resisted Wintel's advance, with Palm Computing[?] being the market leader in palmtops, and Symbian being the market leader in mobile phone operating systems.

See also: Network effect, Monopoly