42 years ago today, on October 28, 1983, Texas Instruments announced it was discontinuing the TI-99/4A computer and withdrawing from the home computer market. It was a stunning admission of defeat for a company that had everything it took to absolutely dominate the home computer market: vertical integration, name recognition, pre-existing relationships with retailers, and a compelling design. Let’s look at what went wrong.
Why everyone was scared of TI
In the 1970s, TI owned the home calculator market. They were the dominant producer of chipsets for small portable calculators. And for a while, TI contented itself selling those chipsets to other companies. Eventually, however, they realized it was more profitable to produce the finished product themselves. So they did. Since they had lower overhead than everyone else, their calculators cost less than everyone else’s.
Almost everyone familiar with the industry expected TI to do the same with home computers. If anything, they could do it faster with home computers, since they could leverage the relationship they had with mass-market electronics retailers.
Texas Instruments’ advantage in the home computer market
TI was one of a very small number of companies who were capable of producing every single integrated circuit required to build a home computer or game console. The others were, theoretically, Fairchild, Motorola, and Intel.
And by everything, I mean everything. TI could make the CPU, the video and sound chips, the memory chips, and all of the 74LS glue logic chips needed to tie all of it together. And TI had all of them in production at the time too.
In theory, TI would have been able to flood the market with an inexpensive home computer and either enjoy the highest profit margins in the industry or undercut everyone else’s price. Once they had all of their manufacturing glitches worked out, potentially they could reach a point where they could do both. And that would have been a nightmare scenario for the industry.
And TI was the only company to do it. Fairchild and Intel weren’t interested. Motorola produced a reference design for a home computer using its chips, but left manufacturing to others. Motorola’s design became the Tandy Color Computer and Dragon 32.
Where it went wrong for the TI-99/4A
TI was supposed to be driving other companies out of business in the fall of 1983, not the other way around. A number of things went wrong for TI, but the the major factors were they didn’t allow open development until it was too late, and their architecture was complicated.
Oh, and an old rival turned up looking for revenge.
How Commodore killed the TI-99/4A
One of TI’s early customers for calculator chips was a typewriter company named Commodore. Commodore had decided to expand into adding machines, which led to making calculators. But then TI decided to start selling calculators, and they started charging Commodore nearly as much for the chips as the wholesale price of a finished calculator.
Commodore’s solution was to try to find a computer chip manufacturer that was struggling, so they could buy them and compete with TI. Commodore found a company called MOS Technology that had recently lost its financial backing. Commodore CEO Jack Tramiel sold his soul to Irving Gould a second time to line up the financing he needed. This later became a problem because Gould effectively gained control of the company and was an incompetent buffoon, but TI didn’t know how much time Commodore had on its clock.
All they knew was Commodore entered the home computer market before they did, and had no intention of losing to TI again. MOS Technology didn’t have the capacity to make all of the components that went into a home computer, and they weren’t especially good at making RAM chips or 74LS logic chips, but they could make everything else.
It was enough that TI couldn’t compete on price. Even though Commodore’s overhead was higher since it couldn’t quite make all of the chips, it made up for it with a simpler design. This meant Commodore’s computers were cheaper and easier to manufacture. Commodore could sell its least expensive computer for $99 and still turn a small profit. TI had little choice but to match the price, but they were losing money by doing so.
The domino effect
And one problem fed into another. The popularity of Commodore’s computer attracted software developers. Commodore had a reference manual that told you everything you needed to know to program the machine successfully. And it only cost $25. While Commodore encouraged third-party development, TI tried to keep it to itself.
When TI decided it was time to open up its platform to third-party development, they didn’t have anything like Commodore’s documentation. And even if they had, their CPU was so unusual, the Commodore was still easier to develop for even if all other things had been equal.
It wasn’t long before Commodore had TI engaged in a two-front war. The $99 VIC-20 undercut TI in price. The $299 Commodore 64 was more powerful than the TI-99/4A and had better profit margins. They weren’t compatible with each other, but both the VIC-20 and C-64 had larger software libraries than the TI-99/4A as well.
TI had no choice but to match the VIC-20’s price and take a loss.
Texas Instruments computers after the TI-99/4A
TI had a successful run selling business PCs later in the decade, and a very successful run selling its Travelmate laptops from 1990 to 1997. But after the TI-99/4A, they never entered the home computer market again.
I still think if TI had either fixed their broken implementation of the Intel 8080 CPU and used that, or used a Z-80 CPU, the TI-99/4A would have been very successful. After all, the rather successful MSX computer, Coleco Vision game console, and Sega Master System game console were all basically a TI-99/4A with a Z-80 CPU.
Today, the TI-99/4A is generally considered a failure. It was a financial failure, to be sure. But it did sell 2.8 million units. So I count it as a commercial success. Problem is, eventually you do have to figure out how to turn a profit.
Winding down the TI-99/4A
Texas Instruments spent the rest of 1983 and the first quarter of 1984 selling down its remaining inventory. Even after announcing its discontinuation, the TI-99/4A sold well. The price was attractive and it had a large enough software library to be interesting, especially after Atarisoft started producing software for it. For people who wanted a computer but were waiting for prices to come down, the TI-99/4A was a unique opportunity in late 1983.
The TI-99/4A retained a large enough install base to have a dedicated following for years as an orphan computer. One of its more unusual peripherals was an expansion option to make it IBM PC compatible.
Once TI sold out of its inventory of computers and peripherals, it still had an inventory of spare parts like keyboards and power supplies. These ended up at surplus electronics dealers, and Radio Shack sold them for a time. Radio Shack marketed the TI keyboard as catalog number 277-1017. Timex-Sinclair owners soon figured out how to rewire the TI keyboard to work with their machines, giving them a nice upgrade over the cheap calculator-style keyboard their machines shipped with from the factory.
I also seem to recall seeing a DIY numeric keypad using the keycaps and switches from a TI keyboard, but I don’t recall if it was for a Commodore or Apple computer.
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.