Good article, quite interesting, especially this tidbit in particular:
>"One other choice that was made early was to use CAD rather than pencil and paper drafting and circuit analysis. This required the company to design their own CAD workstation and processor design software. "
I wonder why/if no other workstation or CAD software was deemed acceptable? It's like Hitchiker's Guide to the Galaxy, they had to build an entirely new computer to the design the Transputer.
I think the author either downplays or is unaware of the magnitude of the T9000 debacle. I worked for people who were very much sold (almost in a cultlike fashion) on INMOS promises for this part, which were never delivered.
Barron was gone by that time so it's out of scope for a Barron-focused article.
I also left before the T9 was finished so I'm not aware of the debacle. I got the impression it was canceled due to lack of potential demand and loss of performance edge as other vendors (Intel) caught up on process performance, with ST unwilling to fund the catch up. I have some T9 die somewhere so it existed in some state at least.
>"One other choice that was made early was to use CAD rather than pencil and paper drafting and circuit analysis. This required the company to design their own CAD workstation and processor design software. "
I wonder why/if no other workstation or CAD software was deemed acceptable? It's like Hitchiker's Guide to the Galaxy, they had to build an entirely new computer to the design the Transputer.