1970s and 80s: Commodore, the MOS 6502 CPU, and Squandering Your Lead

Mismanagement and infighting

Part 1: Rise

The story starts at Motorola: they had just completed their first microprocessor, the 6800. Some of the design team members then wanted to make a simpler, lower cost version, because recurrent feedback from potential customers was that the 6800 was very expensive, and it supported many more computing instructions than they needed. If the 6800 was redesigned to be simpler, supporting only the instructions customers needed, then it could also be made smaller and cheaper. However Motorola management at the time were beset by financial and other problems and weren’t interested in the low cost 6800 proposal. So some of the design team members upped and left, and eventually found another chip maker who accepted their proposal, MOS Technology. So the team finally started to design their simpler, low cost microprocessor, which became the MOS 6502.

In a happy coincidence, MOS Technology was deploying some new techniques in their semiconductor foundry while the 6502 was being developed, techniques that reduced the cost of making a chip to a fraction of what it cost Motorola and Intel. This combined with the smaller, simpler design of the 6502 meant that it first went on sale for 25$, forcing Motorola to drop the price of their 6800 from 175$ to 69$; Intel did a similar price cut. The 6502 sold like hotcakes, and became one of the processors that set off the personal computer revolution, being used in the Apple I, II and III, many Commodore PCs, many Atari PCs and game consoles, and the Nintendo NES.

However, despite the 6502’s success, MOS Technology was having larger financial difficulties: a large part of MOS’s business was selling chips for calculators, but the calculator market had collapsed: Texas Instruments, the biggest calculator chip maker, had started making full calculators themselves, and undercut the rest of the market in price. This market shock hurt MOS, calculator maker Commodore and many others; luckily Commodore was saved by one of their investors, and used the new money to buy some calculator component makers to secure Commodore’s supply; hence Commodore bought MOS Technology shortly after the release of the 6502.

Chuck Peddle of MOS convinced Commodore that calculators were a dead end, and they should make PCs. Earlier Peddle had made a small computer board to demonstrate the 6502, and was now tasked to make it into a whole computer: the result was the Commodore PET, one of the “1997 Trinity” PCs that fueled the personal computer revolution. Commodore later made the VIC-20, Commodore 64 and 128 using the 6502 or improved variants. The Commodore 64 became the best selling PC of all time, after Commodore started a price war with the other PC makers and sold the Commodore 64 for very cheap.

Part 2: Fall

So, Commodore owned the 6502, one of the best selling chips of the day, had a cutting edge semiconductor foundry, and the best selling PC of all time. So what went wrong? In short, they were successful in the 8-bit era [1], but entered the next era too late, releasing the Amiga 1000 four years after the IBM PC. Commodore was eventually squeezed out of the PC market by IBM PC compatibles, like most PC makers of the day.

The biggest reason for this late entry is that they never made a 16-bit successor to the 8-bit 6502. Their focus on low cost helped them gain a lot of market share, but PCs made with the 16-bit Intel 8086 and 8088, such as IBM PC compatibles, were more powerful and were coming down in price. Commodore eventually decided to skip 16-bit altogether and make their next PC with the 32-bit Motorola 68000. By buying CPUs from Motorola rather than making them themselves, Commodore wasted their advantage of having their own CPU design team and fab.

Next came infighting, which lead to the departure of cofounder Jack Tramiel. Tramiel then founded Tramel Technology [sic] and hired away some of Commodore’s engineers, and set them designing a new PC using the 32-bit Motorola 68000. Tramiel then bought Atari’s failing consumer division and renamed his company to Atari Corporation. The new PC was eventually released as the Atari ST.

Meanwhile Commodore acquired a startup called Amiga Corporation, which was designing its own computer based on the Motorola 68000; this was later released by Commodore as the Amiga 1000, and started their Amiga line of PCs. Thus a bloody rivalry started between Atari and Amiga, who were both selling similar systems. The infighting at Commodore had caused valuable talent to leave, and created a fierce rival. But soon both were pushed out of the market as IBM PC compatibles took over; Apple and their Macintosh line were one of the only survivors of the IBM PC invasion.

Part 3: History Repeats Itself

Commodore’s story is common in the tech industry: striking gold once happens for many companies, but few companies can turn that lead into lasting success; when the next industry shift happens, such as shifting away from 8-bit PCs, many companies don’t adapt fast enough and lose their success. Intel, Apple and IBM are rare cases of tech companies that were leaders in their field, then continually adapted and remain successful today.

And MOS Technology’s tale is also unfortunately common: a small company has a successful product, gets acquired by a larger company with different plans for their product, the product line ends soon after. HP killed Palm, Qualcomm shelved Nuvia’s server plans, Intel killed Nervana’s hardware, AMD killed SeaMicro’s microservers; life is tough for a small company.

Footnotes:

[1]: 8-bit, 16-bit etc. is an indicator of the complexity of the CPU: higher bit values usually mean that the CPU is more powerful but more expensive to make, due to the increased complexity.