45 years ago today, IBM introduced the original IBM PC. This computer changed the industry and change the world. There is no shortage of analysis about its impact, but I find a lot of it self-serving and, in some cases gamma a bit inaccurate. In this blog post, I will explore the inspiration of the IBM PC and talk about some of the legends that emerged from its wake. I think it’s a pretty good story. Some of it is even true.
How IBM studied the personal computer industry
In 1980, the personal computer industry was growing quickly. Ushered in by the Altair in 1974 and accelerated by the Apple II, TRS-80, and Commodore PET of 1977, the number of small computers sold in 1980 was approximately double that of 1979. IBM was the largest manufacturer of computers in the world, but the smallest computers they made cost around $20,000 and were designed to support a medium sized business employing a few hundred people. The first iterations of what became AOL originally ran on one of these computers. Yes, one.
IBM wanted part of this growing market. But they’d already seen companies enter it and fail. So they studied the industry to see what worked and what didn’t, and explored various options for entering the market. IBM admired the Atari 800 home computer, and one option they explored was simply licensing that computer, whether that meant having Atari make it for them or make it themselves and pay Atari a royalty.
But in the end IBM took a different route. Atari used a combination of off-the-shelf parts and parts their engineers designed themselves. Texas Instruments entered the market with something entirely their own design, including the CPU.
IBM was certainly capable of taking TI’s approach, building something completely proprietary and owning it all. But instead they came up with something with expansion slots and built entirely of off-the-shelf parts.
How the Apple II influenced the IBM PC
IBM looked at the Apple II and saw how it was more limited than the Atari 800, at least the way it came from the factory. But they saw the proliferation of plug-in cards that expanded the machine’s capability. They wanted to mimic that expandability. IBM also saw how two guys working out of a garage built it from off-the-shelf parts and went from a concept to a sellable product in about a year. They couldn’t take TI’s approach and bring a product to market that quickly.
They also didn’t think the TI approach was worth it. IBM estimated they could sell around 1,000 units in its first six months on the market, and perhaps 25,000 units total. With that kind of volume, it made more sense to build something from off the shelf parts.
I’ve even seen speculation that the Microsoft Softcard, which allowed the Apple II to run CP/M, caused IBM to approach Microsoft to get both Basic and CP/M. The Apple version of CP/M bore a Microsoft copyright, perhaps leading IBM to mistakenly think it was a Microsoft product.
What IBM wanted
After studying the industry, IBM decided they wanted to build a computer around an Intel 8088 processor, put it in a two-tone off-white box with a typewriter-like keyboard with expansion slots, and run the CP/M operating system and Microsoft Basic. CP/M was the de facto standard business operating system for small computers, and it ran on the similar Intel 8080 and Zilog Z-80 processors. The 8088 was the 16-bit successor to those 8-bit processors and designed to be similar. Adapting CP/M software to an Intel 8088-based IBM PC would be straightforward.
IBM didn’t get CP/M, for reasons that are now the stuff of legend, but they got something close enough.
Microsoft Basic, like CP/M, was a de facto standard, an easy to learn programming language that ran on most other computers. Adapting software written in Microsoft Basic for other computers would be easy.
If you want to know why IBM didn’t use a cheaper processor like the MOS 6502, or a more advanced processor like a Motorola 68000, or a more common processor like the Z-80, that’s why. They saw the Intel 8088 as a middle ground CPU with a lot of upside. It was more powerful than the 8080 because it scaled to higher clock speeds and could address more memory, but it was designed to be easy to adapt software written for the 8080 to it, and there was a lot of software for the 8080 available. It was a forward-thinking choice without being too bold.
Open architecture
Much has been said about how the secret of the IBM PC’s success, as well as its undoing, was its open architecture. IBM published a detailed diagram of its expansion bus, as well as the complete source code of its BIOS. This encouraged third parties to produce hardware and software for its new computer.
Where CTRL-ALT-DEL came from
The famous CTRL-ALT-DEL key sequence became a legend in itself. It’s an awkward key combination on modern keyboards, and was even more awkward on the original IBM PC keyboard layout. IBM engineer David Bradley created the feature as a way for developers to soft-reset the machine without having to power off and power on and go through a minutes-long memory test. Bradley has been quoted as saying he invented CTRL-ALT-DEL but Bill Gates made it famous.
Cloning the IBM PC
The IBM PC’s open architecture also meant other companies could clone it. That was part of the reason IBM published its BIOS. By publishing it but claiming copyright, if someone copied the BIOS, IBM could sue them for copyright infringement. If someone tried to reverse engineer it, chances were they’d still violate the copyright and IBM could force them off the market. The idea of software copyrights was still unproven in 1981, but IBM was big enough they didn’t have to win a judgment to win. If they couldn’t win over a judge, they could tie you up in court long enough to bankrupt you.
Admirers of Michael Dell point out that Dell figured out an IBM PC that cost $3,000 contained about $600 worth of parts. But he was hardly the first one to figure that out. By the time Dell figured it out, other companies were making clone motherboards containing a clone BIOS, and all Dell had to do was order them and put the parts together.
Cloning the BIOS was the hard part. Companies like Compaq and Eagle Computer did it themselves using clean-room techniques where one team read the source code and wrote specifications. Another team wrote code from the specifications. Sometimes they even employed programmers who had never developed for an Intel CPU before. But this didn’t scale, and by 1984, companies like Phoenix had developed a BIOS they would license to anyone for $100,000 plus a royalty.
The de facto standard
By adopting and adapting existing de facto standards and lending the credibility of its name, the IBM PC itself became a de facto standard. Various companies quickly came out with their own improved versions of the IBM PC. Notorious examples include the Digital Rainbow 100 and the Tandy 2000. They ran MS-DOS and had Intel processors, but included better graphics, higher-capacity disk drives, and various other improvements. They didn’t try to be 100% compatible, just better, and easily adaptable. All of them proved to be commercial failures. What the public wanted was something completely compatible, or very close to it, and if it was better, that was great, but they weren’t willing to sacrifice compatibility for it.
By some accounts, the IBM PC steamrolled the industry. It greatly exceeded expectations. It took about six months for IBM to sell 60,000 units. And for about 15 years from 2005 to 2020, virtually every computer on the market, including those made by Apple, was derived from the IBM PC architecture.
But how quickly that happened is frequently overstated. The IBM PC and its clones sold around 60 million units in the 1980s. That’s a lot, but that’s about 55 percent of the computers sold in the 1980s. The rest of the industry didn’t just go away overnight. New Z-80-based computers running CP/M continued to be introduced after 1981, and entirely new architectures continued to be introduced into the mid 1980s. But the IBM PC’s momentum built over time, and it came to dominate the industry, even if there wasn’t anything particularly innovative or remarkable about it. It emerged as the cheap enough, good enough solution.
Was the IBM PC’s introduction a success or failure?
This was good for IBM in the short term. In the long term, IBM couldn’t win a race to the bottom. IBM’s success in the 90s was uneven at best, and by the turn of the century, it was no longer turning a profit on its PCs. On December 8, 2004, IBM got out, selling its PC business to Lenovo. In another year and a half, Lenovo will have owned the IBM PC line longer than IBM did.
IBM made some mistakes along the way. The PS/2 line was an infamous example. With the PS/2, IBM tried to squeeze royalties out of its competitors and tried to gain back control after the momentum had shifted toward competitors like Compaq. In the short term, the IBM PC’s introduction was very good for IBM. It far exceeded IBM’s wildest expectations. In the long term, IBM’s mistake was overestimating how much it could control the beast it created. IBM got too greedy, and once it was willing to concede, it struggled to recover the loss of trust.
I don’t think anyone thinks it would have been a good idea for IBM to try to win a race to the bottom. The question is whether IBM could have positioned itself as a premium brand like Apple did. The Thinkpad laptops were a good comeback for IBM. Had IBM repeated the Thinkpad’s success with a desktop machine, it may have stayed in the PC business.
David Farquhar is a computer security professional, entrepreneur, and author. He has written professionally about computers since 1991, so he was writing about retro computers when they were still new. He has been working in IT professionally since 1994 and has specialized in vulnerability management since 2013. He holds Security+ and CISSP certifications. Today he blogs five times a week, mostly about retro computers and retro gaming covering the time period from 1975 to 2000.