- One of the most interesting stories around bubble memory may be that of the IBM Aquarius. Contrary to the common notion that IBM had "totally missed the personal computer revolution of the 1970s", IBM had indeed several personal and/or home computer projects in the works, some internal projects, and there are also sketches of designs commissioned to Eliot Noyes Associates.
The most complete of these projects must have been the IBM Aquarius, as of 1977, which had progressed to fully designed production prototypes, developed under the lead of Bill Lowe and featuring a quite gorgeous, dark-red product design by Tom Hardy, which can be seen in "Delete" by Paul Atkinson (Bloomsbury Academic, London, New York, 2013). The most intriguing aspect of this project probably was that it would come with an entire ecosystem for the retail of its software (think IBM going full Apple App Store in the 1970s), as featured by a card slot on the right side of the computer, where software on cards was to be inserted and could expose special functionality by a (sensor) keypad that would show up in key-size cutouts in the case. And, as may be guessed from context, the software on these cards was to be delivered by the wonders of bubble memory. However, while bubble memory was often touted by IBM as a future technology, it eventually proved as the Achilles heel of the project, when the completed prototype was demonstrated to the executive board and concerns were raised regarding the reliability of the unproven technology.
Tom Hardy, as quoted in "Delete": "The Aquarius would have blown the socks off of everybody. I felt, and a lot of people felt this was going to be a big deal and make IBM believe in this whole business." And, regarding the demise by bubble memory: "Because it was relatively new for these kind of applications and whereas the team had gotten this thing to work, the company just didn't want to take a chance and push it. IBM should have been able to do it. If they would have pushed that technology and put all the resources behind it like they'd done with other things in the past, a lot of folks thought that it would have been successful and would have just blown the whole thing wide open."
While the IBM of the mid-to-late 1970s was evidently not the company to do this (and I can see the risks that might have been posed by this to the reputation of what was still the core product), it's one of the more fascinating what-if scenarios.
[Edit] The IBM Aquarius prototype can be seen here, side-by-side to the earlier "Yellow Bird" (which seems to have been more of a design study than a full-fledged project): https://saccade.com/blog/2019/04/delete-a-design-history-of-.../
- I have questions. E.g., How ist the IBM 5100 a microcomputer? (It doesn't feature a MPU.) Or, the Sharp MZ-80K (1978) is missing (arguably both from the kits and the complete systems lists)…
- As a lesser known fact, the second parameter in "WAIT 6502,1" works as a multiplier. So, "WAIT 6502,2" will print the message twice, with 3 it will print trice, and so on. Even 0 works, filling the entire screen.
- To play around with a virtual PET 2001, see: https://www.masswerk.at/pet/
(Click the button "Prg Library" for some games and more.)
- If you thought that clicking on a word to insert a hyphenation was something invented maybe in the 1980s, well, all you need is light pen.
What about this processor excludes it from being considered ? Asking for a friend …
A PALM processor can be seen here: https://hackaday.com/2023/12/19/bringing-apl-to-the-masses-t.../
[Edit] To be fair, the Wikipedia article on the PALM processor does mention this: "The PALM processor was a circuit board containing 13 bipolar gate arrays packaged in square metal cans, 3 conventional transistor–transistor logic (TTL) ICs in dual in-line packages, and 1 round metal can part."
Compared to pong which is clearly just circuitry
Also. Pentium ii comes on a board / cartridge. Is it not an mpu ? https://en.wikipedia.org/wiki/Pentium_II
The Intel 8008 was actually begun before the 4004 and implemented the design of an earlier, discrete component processor (of the Datapoint 2200 from 1969) as a microprocessor. (This project was initiated by the Computer Terminal Corporation, AKA Datapoint, who comissioned a microprocessor implementations of their processor to Intel, Fairchild and Texas Instruments. The Intel project was the only one to succeed, but the implementation was rejected for being already outdated and too slow at the time, leaving Intel with rights to the design. The rest is history…)
Generally, early coin-op video games, like Computer Space, Gotcha, Space Race, or Pong are not considered microprocessor based, they even lack a CPU in the stricter sense. They are really just somewhat more complex video pattern generators, at least, this is what Nolan Bushnell called them – and he ought to know ;-).
(These early games lack any stepping logic, apart from single-directionally progressing counter chains, which drive discrete TTL logic, which in turn produces the video signal. Movement is exclusively effected by presets to these counter chains and gameplay, like object collisions, is determined by simple logic gates, based on the state of these counter chains. There is nothing Turing-complete in this logic. Commonly, Tank is considered the first coin-op game to use any large scale integrated logic at all (masked ROMs for sprite data instead of discrete diode matrices) and Western Gun, AKA Gun Fight (US-release in 1975) to be the first arcade video game to be based on a microprocessor.)
Regarding the Pentium II – well, that's a tricky question. There are a few contenders to the title of the first microprocessor, even predating the i4004 (e.g., the Four-Phase Systems AL1 from 1969,) the claims of which are generally rejected for not containing the entire processor in a single chip (but two or three of them). Following this logic and tradition, we had to reject the Pentium II, as well. However, due to another, completely unrelated marketing tradition, we commonly do consider the Pentium II to be a MPU.