What was the DEC Type 30 vector graphics display used for?

Question

The DEC PDP-1 had, unusually for its time, a vector graphics display. According to https://en.wikipedia.org/wiki/History_of_Digital_Equipment_Corporation

Most systems were purchased with two peripherals, the Type 30 vector graphics display, and a Soroban Engineering modified IBM Model B Electric typewriter that was used as a printer.

I am surprised that the vector graphics display was that popular. It was famously used for Space War, one of the contenders for the title of first video game, but of course it wouldn't have been purchased for that.

Just what was it used for?

Answer 1

In that era you couldn't afford to build the raster displays that we have now. The RAM for the frame buffer would have been far too expensive. Vector displays were common, even though they had disadvantages - at engineering time you had to make a fixed-for-all-time choice between how long of a vector you could write vs. the persistence time of the phosphor. (Longer persistence time meant you could write a longer vector - but refresh rate was slower, until you could no longer refresh fast enough to allow the user to perceive continuous motion - I'm not even talking about smooth motion, but just the illusion that it was continuous rather than jumping about. Shorter persistence meant you were limited in how much you could write before the user started getting headaches from flicker.)

Terminals at the time only needed 2000 bytes of RAM - at most! - which would hold characters, that indexed into a ROM that had the pixel patterns for the characters. A 640x200 display - CGA - would have needed 16Kb RAM and would have been considered nearly useless compared to the vector terminals which had typically a positioning resolution of 1K or 2K in both X and Y.

So what did customer use vector display for? All the cool stuff you'd expect. CAD (for various engineering domains). Visualization of engineering or scientific simulations. Visualization of the status and performance of operating machinery or installations. Operator consoles for various machinery or installations.

And oddball stuff too ... it just depended if you had access to one or not. E.g., at Harvey Mudd College in 1976 I wrote an IDE for manipulating Mayan glyphs which was used by a professor to prepare diagrams for papers he was publishing on them. (It had a mode where you could draw new glyphs in a vector editor, and another mode where you could arrange the glyphs on a rectanglar grid - with subrectangles - according to the rules of Mayan ... grammar, I guess.) (At HMC you printed your resulting diagrams to this miserable electrostatic "line" printer - it printed high resolution dots just fine but on this stupid and expensive and smelly photosensitive paper so that your print job came out as dark grey on light grey - unless you left it on a windowsill where the sun could get at it which turned the whole thing black).

(Source: Me. In the 80's I worked for a company that designed and built vector-graphics dedicated CAD workstations - Vector Automation, in Baltimore, MD. I also used an IMLAC PDS-1 extensively while in college, which I covered in another answer.)

Answer 2

Just what was it used for?

The answer is in the name - to display graphics

It doesn't make much sense to produce a list of application, but lets look at the core issue of having a screen at all: There were no of-the-shelf graphic terminals - or (CRT-based) terminals - at all. A terminal was a card punch, a tape punch (and their counterparts) or a typing terminal like a FlexoWriter, IBM Model B or C, or any other TTY (*1).

Before vector displays, every output has to be done using either any of the above (hole or text based) devices or a plotter. At that time plotter weren't a comodity perhiperal, but devices with special custom interfaces, usually analogue. They were quite expensive, easy calling 40,000 USD and above. Using them was still a (comparable) slow process taking anywhere from 10 minutes to several hours. Getting to see a result within seconds on a screen was a game changer.

The fact that "most" (*2) were delivered with a Type 30 is simply due the the display being the most singular USP for the PDP-1.

At that time (1959), competition was the IBM 740/780 setup which cost 40,000 USD/year in rent. Just for the Display. A computer had to be rented as well. And even 5 years later, in 1964 the IBM 2250 Graphics Display Unit Started at 280,000 USD (sales price) for a basic configuration (*3).

A 120+ grand PDP-1 with display was already a bargain if only used as terminal to some existing comuter. And with the computer thrown in "for free", it enabled small (*4) departments getting both a once.

---

*1 - See this answer for a partial overview of terminal development.

*2 - At that point it might be important to keep in mind the the phrase "Most systems" sounds bigger than it is, as in total less then 60 PDP-1 were ever build.

*3 - Though, being expandable to 4 displays, this may put a per workstation price at or even below 100,000 USD :)

*4 - Well, relative compared to the ones that could afford an IBM 704 with 740/780. It'll be still an investment worth more than a million USD of today's money.

Answer 3

The PDP-1 was priced to be sold to a lab and was relatively easy to interface to. This means that it could be used for processing and displaying data from a variety of equipment such as mass spectrometers and so on. It was also used by ITEK for one of the very early CAD systems in 1964, the Electronic Drafting Machine (PDF warning).

It could also be used as an adjunct to a larger mainframe computer, a bit like a very expensive graphics terminal which manipulated data that would be produced and/or processed elsewhere. You didn't want to spend time interacting with a computer that you were charged by the minute for. Its price allowed someone to use it directly and it was owned outright rather than leased as many mainframes were in those days.

An early preview of the system in Datamation noted that one proposed application was weapon simulation, perhaps predicting Spacewar?

Answer 4

The PDP-1 was designed as an unofficial successor to the MIT Lincoln Labs TX-0, which also had a similar display. On the TX-0 this had been used for interactive debugging as well as graphical display of the results of experiments (eg an early machine learning experiment that simulated the way a mouse learns to navigate a maze); it is likely that these applications inspired DEC to include a similar display in their machine, and indeed a PDP-1 debugger inspired by the TX-0's system was developed.