16

I'm trying to do research about the Wang VS series, and there's very little on these computers on the internet (as the company was effectively dead before the internet era, this is hardly shocking).

I've stumbled across a scan of a sales brochure that suggests the final machine (VS 18950) had a CPU designated by Wang as the CP-18.

Another (partial?) listing on a hobbyist website suggests the first CPU in this line would have been the CP-3, with most of the numbers (but not all) in between being actual offerings over the years.

Given the era and market (minicomputers), I doubt that this CPU was a single chip, at least initially (and given the high cost of fabbing, maybe even the last).

Are any details known about this chip? Was it a full CPU by modern standards, or lacking hardware stacks and what not like we see on some of the more primitive sorts? Is the instruction set documented?

Improve this question
  • 3
    Welcome to Retrocomputing. Thanks for the detailed question; near-"dead" computers are especially in need of documenting lest the information about them is lost. – wizzwizz4♦ Jan 6 '17 at 18:39
  • I don't know anything about the actual computers, but I do live in Lowell, MA, where the three huge towers that Wang built are still a landmark (some might say eyesore) near the interchange of I-495 and Rt. 3. – Dave Tweed Jan 6 '17 at 20:36
  • Leon Story might be a resource, if you can figure out how to get in touch with him. – Dave Tweed Jan 6 '17 at 20:57
  • Oh I was using a Wang in the internet era, it was 1982-1985. They hid details of the CPU well behind the OS. – Chenmunka♦ Jan 7 '17 at 19:16
  • There's quite a bit of info on bitsavers. – dirkt Jan 7 '17 at 21:01
10

According to the materials I found (I just Googled the phrase "Wang VS series CPUs" from your title), the VS series of processors implemented a slight superset of the IBM System/360 instruction set architecture. Definitely a "full CPU" by any standard. This put them in the class of "high-end minicomputer"; Wang deliberately avoided marketing them as mainframes.

At some point in the early 1980s, IBM and Motorola collaborated on a single-chip implementation of the S/370 ISA, based largely on rewriting the microcode of a 68000. It was used on a plug-in card for the PC, along with a bunch of memory, in order to give it the ability to run mainframe applications. I strongly doubt that Wang would have used this — I mention it only to give an indication of what was possible at the time.

Improve this answer
  • "the VS series of processors implemented a slight superset of the IBM System/360 instruction set architecture" I was under the impression that their instruction set is S/360 compatible at the application level, but not at the OS level – some of the privileged instructions are incompatible with S/360, but that doesn't impact any application, only operating systems. Can anyone else confirm if this is true or false? – Simon Kissane Jan 6 at 9:02
7

This thread seems a little dated but just on the chance that John O is still listening I thought I'd mention that I worked for a major Wang client and I still have a working Wang VS system. Wang tried to avoid using the term "mainframe" in their product lines because they wanted potential users to perceive that the VS series had a much lower overhead cost than a typical mainframe. They did however pitch that the 7000 series were equivalent in capacity and capability to IBM mainframes. I have had experience with both IBM and Wang large systems so I can say with confidence that the Wang VS could do everything imaginable on a large system and, especially in terms of user experience, quite a bit more.

John O: were you ever able to find the information you wanted? I have all the user documents from my personal Wang VS but it won't tell you much about the other models. I am curious to know what you did find. Large Wang systems seem to be pretty rare in the retro community so I imagine finding information could be quite difficult.

Improve this answer
  • 1
    [I know, all of this is OT, still] That's a quite interesting information. Is it still operational? I hope there are plans to keep it that way. As a /360 nerd I'd love to see it and learn about its details. – Raffzahn Jan 5 at 1:21
  • Wow, you have a working Wang VS! I wish there was a freely available emulator. I know TransVirtual Systems (if they are still around) sells a VS emulator, but as far as I am aware they don't have a hobbyist program or anything else like that. I'd love to run a Wang VS emulator just to play with it. – Simon Kissane Jan 6 at 9:08
3

Raffzahn: I have had possession of this Wang VS since new in 1987 and I have every manual, document, advertisement, cable and accessory to go with it. I have two Wang branded 928 terminals and the unique twin wires to go with it too. I still have the terminal and print servers (AST 286 PC with special 16 Bit ISA interface card) and associated software. The terminal/print servers allowed it to talk to devices on an ethernet network. I plan to keep it together until I can transfer it to the custody of a Canadian Computer museum (none currently exists).

Right now the main processor is in my basement and the mock-computer room I built is two floors above it in my house. I am waiting for the Pandemic to lift before I hire movers to carry the Wang VS and an IBM AS/400 I also own up 3 flights of stairs. My wife and I just aren't strong enough to lift them.

Improve this answer
  • :-O Wow! (Former Wang VS programmer here, from 1986 to 2000) – chthon Jan 7 at 9:26
1

John O: The more contemporary Wang VS CPUs employed a writeable control store so that they could load processor core firmware at the time of IPL. They came standard with software to load an IBM 360 compatible instruction set because this gave Wang access to a larger pool of talented programmers. They were not IBM compatible in any respect and the rest of the system architecture and the operating system were a complete departure from anything offered by IBM. They were optimized for a fast online user interaction and didn't really provide much support for batch processing.

An Wang, the founder, was a former IBM employee and a substantial IBM shareholder who didn't like the way things were done at IBM so he designed his systems to be completely different.

We did all our programming in high level languages so I don't think I have any assembler documentation. I'll go through what I have to see if anything would be useful.

On another issue:

I noticed that your table 3xxx - Technology companies & businesses doesn't have an entry for Cincinnati Milacron, a process automation and industrial robot manufacturer who for a period of time manufactured IBM source code compatible data processing systems. It was described to customers as an IBM compatible (targeting the System/3) but it had an 18 bit word so I don't think it was a clone of any IBM CPU. I wrote some software for them at a data processing VAR in the early 1980s. The only programming language we had was RPG II and it could execute IBM standard JCL. It supported 1 VDT workstation and our applications were all batch oriented.

Rob Carnegie

Improve this answer
1

Simon Kissane: I have tried to contact TransVirtual Systems to see if they would consider some kind of Retro-Computing offering but I've never even had a reply. I suspect the company may no longer exist. In any case, they probably could only sell the product to somebody that already owns a license to the operating system and development tools etc as I assume they are only emulating the hardware environment.

I think there is was working Wang VS80 at a museum in Rhode Island, it is about 5 year earlier model than my VS 6E.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.