The first parallel processing chip
Leon
Posts: 7,620
Here is an interesting account of how the first parallel processing chip, the Inmos transputer, was designed over 25 years ago:
http://www.cs.man.ac.uk/CCS/res/res32.htm#c
http://www.cs.man.ac.uk/CCS/res/res33.htm#c
I paid £400 each for a couple of T414 chips, at the time! I wire-wrapped a prototype board, with 256k of DRAM; somewhat to my surprise, it worked first time.
http://www.cs.man.ac.uk/CCS/res/res32.htm#c
http://www.cs.man.ac.uk/CCS/res/res33.htm#c
I paid £400 each for a couple of T414 chips, at the time! I wire-wrapped a prototype board, with 256k of DRAM; somewhat to my surprise, it worked first time.
Comments
Most interesting would be in a Propeller chip could be adapted to emulate the T414 for the sake of history. Does such a project interest you?
Still can't get interested by it's so-called successor. It seems to be a throw back to the 1980's with it's memory limitations, then there's the weird I/O.
Yep, I suffer from that as well:)
However, Leon is not out of order to bring up such things here.
The Propeller is after all a multi-core processor that encourages thinking about parallel program execution. As an MCU without any inbuilt peripheral hardware, UARTS, SPI, USB etc etc, it also promotes the idea of creating those, normally hardware, functions in software. That is "software defined silicon".
Just so happens that these ideas are exactly what was in mind with the Transputer chip and Occam language a few decades back and being continued with "the chip that shall remain nameless" now a days.
http://www.atarimuseum.com/computers/16bits/transputer.html
DJ