Norsk Data computers at the Norwegian Telecommunications Museum

Here are some pictures I have taken at the Norwegian Museum of Telecommunications. In addition to a wonderful and well-kept collection of telecommunications equipment - from the earliest manual switchboards, via Strowgers, and everything else up to DSLAMs, they agreed to take a selection of computers Norsk Data had saved away through the ages under their wings - as Norsk Data's service department became a part of Telenor, they fortunately felt something of a responsibility.

The service division of Norsk Data was separated into ND ServiceTeam in 1989, following the stock exchange crash of '87. Later, they were renamed to Comma, later Comma Data Service (which I believe is a rather fitting name, considering the original, literal Greek meaning of the word comma - "piece which is cut off".) This was later bought by Telenor for, I believe, tax reasons. This then became Telenor Comma. As soon as the quarantine time was over, it became a part of Telenor Bedrift (Loosely translated, "Telenor Corporate Services"). The Norsk Data service department was then significantly downsized.

This later becomes a part of Telenor Bravida. Around 2000, Bravida announces they will focus on computing, networks, and UNIX servers. The subsequent day, half the UNIX department is fired... In 2001, the profitable Norsk Data department is suddenly shut down, to the great surprise of both customers and employees. In 2003, the warehouses and workshops are emptied with complete disregard of the rather obvious historical value of the contents. It is extremely frustrating to me to know this now - as by 2003, I had certainly heard the computer history calling, and if I'd been aware, would have helped with saving these machines.

Fortunately, a few people realized long ago that this simply was wrong! The fascinating relics of a company which in 20 years grew - faster than Apple Computer! - from a three-man corporation to a multinational company employing over 4000, second only in stock value to Norsk Hydro - were not wiped out. This is as far as I can tell mainly thanks to the efforts of two men from ND, Jonny Oddene and Kåre Trøim. The Telecommunications museum graciously accepts hosting these computers, despite the relatively loose connection between the history of Televerket and Norsk Data. Thanks to these efforts, these fascinating machines will be around for posterity to study and enjoy.

Recently, I, with a lot of help from the ARPAnet pioneer Pål Spilling, saved a NORD-1 (Serial number 4) and a NORD-10 (Serial number 5) which had been standing in a barn in a rural area in Gjerdrum since the bankruptcy in 1992. I am in the process of restoring the NORD-10. To this end, I needed a board extender, to be able to perform measurements. I was loosely aware of some Norsk Data machines having found a home at the Telemuseum. However, I was not at all prepared for anything even close to this scale!

Originally, I was planning to archive the floppy diskettes on a more recent media - thus me bringing up the test machine - for accessibility and to prevent bit rot, but as the media is considered to be in no immediate danger of unreadability, it's been decided that we postpone this project.

In the long term, I hope to be able to create a thriving, playful, and educational hacker community around these machines, hopefully resulting in increased awareness and enjoyment of these wonderful computers - by making available software, documentation, and providing access to running systems.

If you have any questions or comments, or would even like to see these machines in real life, please contact me.

[big] [huge]

p1090046.jpg

The computer shelf, as seen from center, looking left. To the right, you can spot an ND Butterfly resting atop a Data Products punch, itself resting atop a Plessey ferrite core. Further inwards is what appears to be a more or less complete SAM 2 AKA FLINK 1.

Further inwards, you can spot a NORD-1, a NORD-10(/s?), two late-model NORD-1s, some unknown wiring system, another NORD-10, and one or more NORD-5 computers spread across multiple racks.

[big] [huge]

p1090047.jpg

Now looking to the right from the same point. To the left you see the same Butterfly stack. going inwards, an ND-5800, the machine currently in use as a restoration machine - see the later pictures for more. also, a NORD-10/S with a magtape drive (which for some reason is facing the other way), a very late-model tpServer A45, and an empty 5000-series Filestore rack.

Further inwards, you see an array of Compact and Satellite machines on top, with a DECwriter IV and some old disk storage below. Further inwards is a few more orange-era ND systems, and at the very end is a UNIVAC 9400 which I sadly failed to get any usable pictures of. On the very top shelf you may spot some NORD-100-era disk drives.

[big] [huge]

p1090049.jpg

An overview from the end of the shelving. On the right is some fascinating telco equipment, and on the left a better view is offered of some of the ND equipment, including the NORD-5(s). A NORD-10/S is on the second shelf, and interestingly, an IBM 3274 establishment controller (is that the name?) is the odd one out in the lower left cardboard box (the one with the black top.)

[big] [huge]

p1090058.jpg

Here is just a NORD-10/S picture I tried and failed to get right, for the Wikipedia artcle. I'll try again later, with a tripod.

[big] [huge]

p1090062.jpg

Another failed picture which I included by virtue of nothing really like it existing. I'll Try Harder.

[big] [huge]

p1090066.jpg

Here it is. NORD-10.1! I asked Wiggo to lift it out of the shelf so I could just get a nice look at it. The machine is in very nice condition, but its prototype status is abundantly clear. The NORD-50 next to it may also be a very early one - it certainly is a beautiful machine. The sheer amount of boards is impressive!

[big] [huge]

p1090073.jpg

Here is what I enjoy pretending is my office. :) A nice hideaway to work in.

[big] [huge]

p1090080.jpg

This is the original ND software library. This looks smaller than it is (amazingly). Keep in mind that apart from the first shelf, all the floppies are 8", and the case front covers in the background, leaning on the egg-white shelves, are around 1 meter 50 tall. It seems odd to me that all the way up to 1989, they kept their master copies on 8" floppies!

[big] [huge]

p1090082.jpg

The floppies are stacked quite deep in as well. They are all in excellent cosmetic condition.

[big] [huge]

p1090083.jpg

The boxes with the dates are master copies. When a new version or revision was released, a master copy would be made and put in a dated box. This part of the collection is thus not by product number, but by release date.

[big] [huge]

p1090085.jpg

The PCB films, used to produce the boards themselves. They seem to have films for every single PCB put in production by Norsk Data. Best of all, these films, as the black and white photos (I assume) they are, will last for centuries and not need immediate attention. This particular board is the PCB used to connect SCSI devices inside a late-model (gray) "Comson" (Same size as the case known as "David" - the brown Compact case, I believe.) Compact chassis.

[big] [huge]

p1090090.jpg

One of the three red cabinets full of source media, sorted by product number.

[big] [huge]

p1090093.jpg

An example of how a source original looks. This is the proprietary email system NOTIS-ID

[big] [huge]

p1090094.jpg

I've taken some pictures of the folders to provide some examples of what's in here. here, we see ND-500 CPU diagnostics and User Environment. D03 was a relatively late release of UE, relatively late into the 1980s.

[big] [huge]

p1090097.jpg

Technovision was an excellent modular CAD/CAM system which was internationally considered to be one of the best on the market. It was in part (I do not know how big a part) designed by Norsk Data Dietz GmbH. A special workstation named the Technostation was designed specifically for running Technovision. It was extremely well recieved by international press, and even won a design award. (warning: Huge picture, and Norwegian to boot.) I would not personally mind playing around with one some time. :)

[big] [huge]

p1090100.jpg

NOTIS-ENCRYPT made me very interested. I'm curious to see what kind of encryption methods they chose to implement on a comparatively small 16-bit CPU. (Note to self: It says "for ND-110". What extra instructions were added to the 110? Check the product announcement, I know I have it.)

[big] [huge]

p1090101.jpg

A complete OSI stack for SINTRAN III. I don't know whether to laugh or cry, but it would be fun to poke around with a bit.

[big] [huge]

p1090109.jpg

A perspective look at the source cabinets. It really is quite a lot of source code. I'd guesstimate at least 200-300MiB of source code. This used to be the bread and butter of a 4000 employee strong company!

[big] [huge]

p1090112.jpg

The system information sticker from the ND-5800 machine I've chosen to use as my test machine. The 5000 CPU inside it is however an ND-5700.

[big] [huge]

p1090118.jpg

A nice image I took while playing around with extreme closeup mode on my new camera. A bit blurry, unfortunately. If I get a hold of a tripod, again, I think it could become a very nice picture.

[big] [huge]

p1090123.jpg

Coffee, Classical music, and Comprehensive documentation!

[big] [huge]

p1090124.jpg

This potent-looking trio of PSUs can supply an impressive 320A at 5V. The AC pulled by the ventilation system itself cannot be negligible. The airflow is very impressive.

The DC300, which is the support battery-backup PSU, showed a Transient warning (more than +/- 10% voltage off) which refused to reset. I switched in the DC300 PSU module from the tpServer (the gray 5000). This worked. This is why the DC300 depicted here is Seem and not Philips like the rest.

[big] [huge]

p1090127.jpg

I got the machine working two-three minutes before I had to leave. Thus I didn't pay close enough to the trophy pictures I took of it. This one was the least awful one. The lower rightmost board which you can see is not inserted fully, is a ND-120/CX 4MB CPU. I had to swap in the ND-120/CX CPU from the tpServer. This, along with the PSU, are the only swaps I performed - and I have documented these in my paper worklog.

[big] [huge]

p1090134.jpg

And the Test Program Executive, a minimalist monitor for other diagnostics utilities on the same floppy, is running. I did not have time to run any of the (long and exhaustive) diags. But the mere fact that it is loading from floppy, and displaying the TPE prompt does really go a very long way.

I also want to get a Tymshare line printer going, not just for the vanity and fun, but also so I could have a printed record of the diagnostics (a built-in provision), and printouts of my source code.

Valid XHTML 1.1