When it comes to Atari ST vs Amiga, there are more similarities than differences from today’s perspective. But the two machines had significant differences that led them to be incompatible even though the hardware differences look minor today. Here’s a look at the two machines and why they were such fierce rivals in the late 1980s.
The engineer swap
Due to a series of events, a large number of ex-Atari engineers ended up at Commodore designing the Amiga while a large number of Commodore engineers ended up at Atari designing the ST. Jay Miner, the principal designer of the Atari 800 computer, led the Amiga design. Meanwhile, Shiraz Shivji, one of the main designers of the Commodore 64, led the Atari ST design.
In some ways, it felt like the legitimate successors of two of the most influential 8-bit computers of the 1980s ended up being traded for one another. The ST resembled the C-64 more than it resembled the 800, and the Amiga resembled the 800 more than it resembled the 64. The Amiga was the Atariest Commodore ever, and the ST was the Commodoriest Atari ever.
How it happened
In January 1984, Commodore founder Jack Tramiel abruptly quit. Large numbers of Commodore employees left soon after Tramiel, under mixed circumstances. On July 1, 1984, less than six months later, Tramiel and his sons bought the consumer division of Atari from Warner Communications. Tramiel quickly hired numerous ex-Commodore engineers, including Shivji.
Meanwhile, Atari had been bleeding engineers for years. Jay Miner left in February 1979 after a disagreement with new CEO Ray Kassar. Joe Decuir, another key engineer, left soon after in June. In 1982, Miner founded a small company called Hi-Toro, which after a few name changes, became Amiga. Miner quickly brought in Decuir as a consultant to help with the design. Amiga’s main product was a next-generation game console, but when the bottom fell out of the video game market, Miner and his engineers simply repurposed their chipset and built a computer around it instead. Meanwhile, Amiga developed and sold a few games and controllers for other consoles to provide some cover and much-needed income.
By 1984, Amiga was running out of money and looking for options. Numerous companies like HP and SGI wanted the chips, but not the rest of the computer. They also talked to Jack Tramiel, briefly, and even to Steve Jobs. The talks with Tramiel went poorly, and contrary to some accounts online, happened independently of the talks with Atari. Tramiel didn’t own Atari yet.
Atari provided the least-bad option. Amiga secured a short-term $500,000 loan from Atari in March 1984, just before the Tramiel purchase. Part of the agreement would allow Atari to use the Amiga chips in a new computer, which would be named the Atari 1850XLD. Amiga backed out of the agreement at the end of June when enough money to repay the loan, plus interest, and to terminate the contract turned up. The benefactor was Commodore, who provided the money to buy time to negotiate a win-win deal. Then in August 1984, Commodore purchased Amiga for $27 million.
The result was by August 1984, key engineers from both companies had changed sides and the two companies had traded lawsuits.
Legal action between Commodore and Atari
Atari sued Amiga for $100 million in August 1984, citing breach of contract. Had the suit been successful, the Amiga technology could have ended up back at Atari.
Meanwhile, in June 1984, Commodore sued over Shivji’s departure along with three other Commodore engineers, saying they took design documents related to the Commodore 900 computer with them when they left. The Commodore 900 was originally intended to be a 16-bit computer based on the Zilog Z8000 running Coherent, a clone of the Unix operating system.
Commodore produced working prototypes of the Z8000-based C900, a few of which survive today. But Shivji headed a skunkworks project to change the C900 to use the Motorola 68000 processor, which was slightly faster than the Z8000. Shivji and his team never got a 68000-based C900 working while at Commodore. But within six months, in January 1985, Atari had a functioning 68000-based prototype computer to display at CES, dubbed the Atari ST.
Both suits ended up being settled out of court and very little detail made it to the public. Atari released the ST in June 1985 and Commodore released the Amiga 1000 in November. Had things gone differently, we could have ended up with the Atari 1850XLD competing with the Commodore 900 instead.
Atari ST vs Amiga hardware
Both the Atari ST and Amiga series were based on the Motorola 68000 CPU or one of its successors. But that didn’t automatically make the two machines compatible. They weren’t. They used different supporting chips and mapped the chips differently internally. Today we can compensate for those kinds of differences using device drivers and hardware abstraction, but that adds overhead that no one could afford in 1985, with 7-8 MHz worth of CPU and 256K or 512K worth of RAM to work with.
Atari ran its CPU at 8 MHz, while Commodore ran it slightly slower, at 7.16 MHz, as this speed was convenient for synchronizing with NTSC video. Atari provided 512K of RAM in its base model, while Commodore shipped the Amiga with 256K standard. The Amiga chips were more complex and expensive to produce, so Commodore couldn’t match Atari on price initially.
On paper, the ST was the better value, delivering 14% more CPU speed and double the memory for almost $500 less. But the Amiga’s chipset acted as coprocessors, meaning the CPU could hand them a job and point them at a bit of memory and they’d go do a task while the CPU was free to work on something else. So as long as there was any type of display going on or sound playing, the Amiga was actually faster. The difference would be similar to two modern PCs, but one of them using onboard graphics with a faster processor and one using a discrete GPU and a slightly slower CPU.
Atari ST vs Amiga operating system
Both Atari and Commodore wanted a Mac-like graphical operating system that used a mouse. Neither company had the resources to deliver it on their own and release it along with new computer hardware in 1985. Atari talked to Microsoft and Digital Research. Microsoft couldn’t port Windows to the ST before 1986, so Atari went with Digital Research. Commodore talked with Berkeley Softworks, the makers of GEOS for the Commodore 64, and with Metacomco. Commodore used a combination of Metacomco’s technology and its own.
Digital Research had a multitasking operating system in the works, but ended up building something much more MS-DOS-like for the ST, and letting its GEM graphical environment run on top of that. The computer hit the market later than Tramiel would have liked. But this arrangement still allowed it to beat the Amiga to market. True multitasking didn’t come to the ST until 1993, with the release of MultiTOS.
The Amiga’s operating system didn’t look quite as nice as GEM, especially at first, but it multitasked from day one. When both companies demonstrated their machines on The Computer Chronicles TV show in 1985, co-host Gary Kildall was visibly interested in and impressed by both machines, but his reaction to the Amiga’s multitasking suggests it surprised him.
Who used the Atari ST and Amiga?
Neither the Atari ST nor the Amiga ended up being as successful in the end as the Macintosh. Initially the ST outsold the Amiga but after Commodore released its less-expensive Amiga 500, the Amiga overtook it in sales. The Mac outsold both of them 2:1, however.
The ST remained a good value play, as it was the cheapest 68000-based computer on the market. As such, it attracted numerous software developers. But developers’ enthusiasm waned over the years due to software piracy among ST owners. The killer app for the ST ended up being music, because the ST had a built-in MIDI port.
The Amiga had similar issues attracting and retaining third-party development. Initially it was popular with graphic designers and for software development, as developers could produce graphics on the Amiga and then port them to other platforms. The Amiga 500 was a popular home computer for a few years. Commodore initially hoped it could replicate the C-64’s success, but it fell far short of that goal.
In the end the killer app for the Amiga was video production. The Video Toaster was heralded as a production studio in a box when it was released, allowing an Amiga to replace other equipment that cost tens of thousands of dollars.
Neither product succeeded in holding onto more than about 3 percent of the market, however. By mid-decade, Commodore was out of business and Atari had merged with JTS, a then-promising manufacturer of hard drives.
So who won Atari ST vs Amiga? Apple and Microsoft. Once their platforms gained color and sound capabilities, big-name software developers put more of their energy into those platforms than either Atari or Commodore.