Transputer

I took a course in Occam at university, though we weren't allowed to program the Transputers themselves. Occam had the nice property of being able to produce code that could be proven mathematically correct. We were allowed to see the output from a Meiko Computing Surface (with something like 64 Transputers) generating fractal zooms in real time (circa 1990).

Several years later I discovered a Transputer (in its storage case) abandoned in the drawer of the desk I was assigned in my first "real" job, so I kept it as a souvenir.

https://en.wikipedia.org/wiki/Occam_(programming_language) https://en.wikipedia.org/wiki/Meiko_Scientific

I learned about the Transputer from Byte Magazine, probably by reading one of Dick Pountain's articles. I thought it was interesting enough that I wrote Inmos to see if I could get some free documentation. I think I must have filled out a standard form, because I remember that they wanted to know what kind of professional role I had. I just wrote "student" and they sent me a bunch of stuff anyway, but the paperwork that came back said I was a "consultant", which I found rather amusing.

Alas, Transputers were on the way out at that point (probably they were never on the way in), so I never got to actually work with one.

They still seem pretty cool, but in a decidedly retro way now.

If you're interested in this, http://www.zachtronics.com/tis-100/ is a "game" about programming a non-existent architecture that looks inspired by transputers.

As a student (around 1990) at DTU (Denmark) a friend and I did a project on trying different methods of mapping backpropagation neural network algorithms/architectures to a computer with 16 T-800's using the Occam language with its (in)famous folding editor. Those were the days :-)

I have attended lectures by David May and he spoke a couple of times about the Transputer. He has some resources on his current web page.

https://www.cs.bris.ac.uk/~dave/

I so wanted to program one, but never got my hands on the Atari or the PC card with the T-400 on it. I bought the Occam book (which is I still have), and found it to be a very interesting language (I always got the impressions on hooking Legos together). I will say that it did influence my thinking on how to design programs.

After re-reading the history of the Transputer, I wonder if instead of going all in on the T9000, maybe Inmos should have been simply trying to cram multiple T800 cores onto a single chip. I guess it likely wouldn't have made much difference.

The transputer still lives in on Xmos (started by David May), http://www.xmos.com/products/silicon/xcore-200

Our visual systems (flight simulators) lab had a small card cage from Meiko. It held a number of Transputer boards and was used for graphics/rendering algorithm development. I remember a flurry of activity to learn Occam, but interest waned as Transputer development fizzled out. I was an apprentice at the time and didn't get to play with the Meiko box. Other dev work was done on PCs, a VAX 11/750 and a MicroVAX - and I got to play with all those!