All those old systems either ran on DEC or Data General machines--PDP/Vaxes or Novas. The Sun workstation-class machines came a bit later. AutoCAD was the first PC-based option, but it was not up to the level of Intergraph or similar at least until the 90's.
The Intergraph IGDS system I used had a wired cursor with a round glass window that had high-quality etched crosshairs. You could see the copper-wire coil sandwiched in the glass. The table itself was the other half of that, and it was a good 40x60" table that contained an embedded high-resolution grid with dual Tektronix storage-tube monitors hanging above the work surface. I think those puppies were about $125K in 1980 dollars for each workstation. I doubt anybody got their money's worth before they were technologically obsolete, so you know companies kept them past their prime. I think Intergraph was the first system for manufacturing that put a dent in Auto-Trol, but we used it for civil-engineering stuff that relied on being able to align geospatial data with real precision. Auto-Trol never did that.
Working on the boards a few years earlier at a company that designed and manufactured conveyor-belt systems showed me that I could do more real design as a drafter in the engineering world than as a degreed architect. It also taught me how to draft for steel fabrication. That's when I switched to engineering in college. But change happens slow in a drafting room--the supervisor required the use of Smoley's Four Combined Tables for trig calculations (and we made hundreds of those every day), because he couldn't figure out how to use scientific calculators. The owners of the company finally put a stop to that, but I defied those orders from the start, being a smart-a$$ college boy. It is true that good scientific calculators were only a few years old at the time. The engineers could use their Keuffel & Esser slide rules, but in the drafting room, we needed to dimension to sometimes five or six significant figures to be accurate enough that the conveyor mounting bolts would align at both ends of several hundred feet of inclined structure. Smoley's just went to five significant figures and was too slow to use at that.
I wish I could remember all that trig these days without having a cheat sheet.
Rick "remembers starting the boot sequence of Data General Nova 1200s using octal register switches" Denney