PDP-10 Digital Equipment Corporation KA10 CPU

[@thomasw.eggers4303]

I helped debug the first KA10 in the Maynard mill about 1968. Alan Kotok chief engineer, Dave Gross, Bob Clements. We had the logic modules, but they didn't have any handles yet, so we went around to manufacturing, found the right machine, put the handles on the modules, plugged them into the back plane, and turned the power on. Nothing seemed to be burning, so started debugging the logic. The OS was running one week later. I was the Field Service engineer on a KA10 I installed at UPenn in Philadelpha. (It was a medical research building.) Got down there and was waiting for the truck to arrive. Instead I got a telephone call saying the truck had driven under a low railroad bridge just outside of Maynard. The truck's top was mostly sheared off, and all the computer equipment flew forward as the truck came to a halt. Some poor guy had his household furniture in the front of the truck, which cushioned the computer equipment but left the furniture in splinters. The equipment was taken back to Maynard and checked out thoroughly again before delivery to UPenn. When the equipment was all up and working, to the computer center's satisfaction, I was handed a check for $650K (and that's 1968 dollars). I was a bit surprised, but folded the check in half, put it in my wallet, and flew back to Boston, driving to Maynard. Handed the check to the DEC VP (Win Hindle), who said, "Thank you," with no further comment. That was, and still is, the largest amount of money I've ever had in my possession.

[@gregoryfinn2461]

Memories ... Brandeis Univ late 1971. Hell of a lot of fun. When that KA was moved into the new Feldberg computer building, care was taken to hook up the memory boxes in the same order. Fingers crossed, the machine was turned on. Tops-10 continued where it left off, pulling register contents from their core locations when halted. But the daytime clock was wrong. Darn!

[@jamesk8923]

I love that story! Did you work for Digital? Jim

[@averydesignvideo]

back in the day KA10 SN7 and SN27 were my DEC field service assignments at Fermi Labs, SN27 was the site timesharing machine and SN7 belonged to a bubble/spark chamber neutrino experiment and ran a custom built triple flying spot film scanner for spark chamber data acquisition along with being fed by a PDP9 that ran a room of manual bubble chamber film scanning systems.

[@jamesk8923]

Hi Dave, You sound like my kind of LCG guy! It was heartbreaking seeing that last surviving KA10 in useful production scrapped. In fact, the guy in the video quit his job at Penn State over it. I still have the bay 1 backplane preserved with all the modules. I worked at Digital from 1969 to 1997 in Maynard, Westminster, Princeton and finally Wilkes-Barre. I know it’s a strange career path, but it’s my hometown.

[@lesnyd]

I began dual CPU KA10s 274/276 in a slave/master configuration in 1974. Had a great time, eventually brought a trip-SMP system on 1090s in 1983. Brings back fond memories!

[@DavidMillsom]

Thanks for posting this. I worked on a KA-10 in Melbourne in the 1970s. It was a great machine and TOPS-10 was very advanced.

[@davidmillsom1474]

@@citizenk3360 Yes, in the mid 1970s. Noel Fenton was an engineer there. It was a great machine.

[@DavidMillsom]

@@citizenk3360 Yes. Noel Fenton was engineering manager. It was a great old machine.

[@DavidMillsom]

I recall that when the Stock Exchange took over maintenance from DEC, they insisted that it be brought up to par. That meant making sure every light on every cabinet had to be operational !!

[@DavidMillsom]

@@citizenk3360 If that's the swapping system I recall, they used the same units in the B5500 (Burroughs 5500) at Monash. When the B5500 was dismantled, my father and I took a couple of those platters home. They were about a 1/4 in. thick and 2 1/2 -- 3 feet in diameter. We thought they'd be great for BBQ but we were never sure whether the coating was toxic.

[@DavidMillsom]

@@citizenk3360 The first machine I learned computing on was a PDP-9 !! Later when I worked at the Stanford Linear Accelerator Center I discovered that one had been used to control the accelerator. It had not been powered on for years and when I did power it up the bootstrap was still there. (At 17637).

[@Edomaden]

This feels like the start of an ARG.... ....Impressive system!

[@lesnyd]

Anyone remember how the Colorado School of Mines set up the null jobe to display "CSM" in the lights?

[@robdoyle161]

Thanks for posting that! I could have sworn that the operator looked familiar - but the layout of the computer room was all wrong. When the video showed that CAD program, I knew that was the KA-10 from the Penn State HCL. I was told that there was only ever one KA-10 that had a 340 display. When I used that machine, the 340 Display had a light pen, not a mouse. I still have schematics that were generated using that CAD program. I have enjoyed almost a 40 year career in electrical engineering, and a large part of that was influenced by the time I spent sitting near that very computer. It saddens me to know that it was scrapped. Such a pity.

[@jrstf]

My mouse interface schematics are dated May 7, 1985. Presumably it become operational near that time. The mouse did not replace the pen, both pen and mouse were operational, but the mouse was much easier to use. The pen only detects light so the CAD program had a "display grid" option to give it something to detect. Displaying a 80x64 grid slowed down an already slow refresh, for a big schematic the refresh rate could be seconds, it was long persistance phospher, but not that long! For the grid, I used the 340 scope's vector scale mode where it only drew every 8th point, therefore 8x faster. The 340 drew points at a maximum rate of 1.5uS per point. The machine was moved at one time, down to the end of the hall, and the EAI680 analog computer connected to it was eliminated, that's probably why it looks different.

[@larsbrinkhoff]

The MIT AI lab PDP-10 also had a 340 display, admittedly shared with a PDP-6 on a common I/O bus.

[@TheNightquaker]

This is one badass piece of kit.

[@oldschoolnomad813]

Awesome!! the first computer I worked with was the PDP-10 at Albuquerque public schools CEC magnet campus from 78-82. I learned programming in BASIC, Fortran, Algol, Pascal and COBOL and Assembly on it. That PDP-10 was also the one that MITS rented time on it for Paul Allen and Bill Gates to develop Altair BASIC. I actually hung out with 1 kid that hacked into the drive that Altair BASIC was stored on what he did with it I have no clue as we went on our separate ways after high school.

[@RS-ls7mm]

Me too. By any chance were you also part of the New Mexico Computer Society? I do remember finding the early microsoft directories. Security was not too tight back then. I remember how crazy fast the line printer was, basically shooting the paper into the air.

[@alexm.6133]

Wow.... upgraded with red LED lights!!! Top shelf... great vid, thanks Wild Man!

[@jrstf]

DEC field service would replace the incandescent bulbs with LED bulbs a few at a time, when the need arouse. That was manpower intensive and the LEDs were pricey, something like $15 each is what I recall. So when I took over maintenance of this machine, I bought lots of standard LEDs and resistors, cut leads to length, soldered a resistor to each leg of the LED. The assembly fit the socket. Worked great, took longer but a whole lot cheaper. The KA10 had 2 or 3 RP03 and 1 RP02 disk drives. Those were 50MB (25MB for RP02) removable pack drives. I had a fear of disk heads, afraid of a crash, afraid to calibrate them. So I replaced the drives with a VAX 11/780 (a general use machine, not dedicated to this purpose). Looks like that happened shortly after midnight on 27-APR-1991. I still have the disk container files off the VAX. Just waiting for a KA10 (and 340 scope) simulator so the KA10 can resume running where it left off. File FACT.SYS[1,4] seems to be the newest file, it is dated Sunday 6-Feb-1994 06:57. So I believe that's when it was shutdown. Perhaps Jim can confirm it was a Sunday (I wasn't there that day).

[@larsbrinkhoff]

Richard Cornwell's KA10 simulator does run TOPS-10.

[@larsbrinkhoff]

Hello Tim, I tried to send a message to Bryan about the KA10 emulator, but I didn't get a reply. Do you know how to contact him?

[@alexm.6133]

In the early incandescent days we'd flip down the cab bezel over the lights and rest our chin on it while pushing the lamp test button... jokingly calling it the best tanning available in a computer room! Not so much with the LEDs... sigh... advancing technology was a buzzkill!

[@larsbrinkhoff]

Thanks for replying. I got hold of Bryan through Facebook (at least it's good for something). I believe he got his TOPS-10 images up and running in an emulator!

[@chotaire]

How must it feel like for Bryan to lose the only remaining PDP-10 that he has spent so much time with, knowing that his code will not run anywhere else forever. How could they scrap this machine? It was the last one. Has anyone contacted Bryan about the new one running at Living Computer Museum? I can imagine this feels like "The Internet has now been turned off, do something else from now on" for us network engineers.

[@DVRC]

So, all PDP-10 are gone?

[@GodEmperorSuperStar]

He can run his code in SIMH, KLH-10, etc.

[@marcoantoniotadeudasilva7144]

Wonderful

[@vcolino]

Amazing!

[@citizenk3360]

Thanks Juan! I loved working on the PDP10.

[@vcolino]

@@citizenk3360 What a lucky guy! :)

[@nzoomed]

Did these run on integrated circuits, or just pure transistors?

[@jamesk8923]

There were a few chips in the “fast accumulators” option in bay 2 but all the other roughly 1500 modules had transistors.

[@LeeCourtney]

Is that a box fan stuck in a frame at 5:57? 😼

[@VulcanOnWheels]

I'm so sorry the video is not clearer than this.

[@citizenk3360]

The original handheld 8MM video is clearer, but not as steady. The KZbin editor calmed it down but blurred it quite a bit. I still have the original video.

[@redmartian]

@@citizenk3360 I think there could be a chance for improvement. I'd be happy to help.

[@guywithillegalname]

Do you recall when the video was actually shot? Or am I just missing it in the text there?

[@guywithillegalname]

James Kinsman close enough for my curiosity, anyway. Thanks!

[@larsbrinkhoff]

HEY! We have a KA10 booting again! kzbin.info/www/bejne/onSalJ9rptaVm6M

[@BrokebackBob]

Where is this beauty? Is it running TOPS-10 Timesharing Operating System?

[@BrokebackBob]

James Kinsman In college at Indiana University Bloomington, IN main campus, I learned at least 5 programming languages via timesharing on a monster DECsystem-10 KL10. It changed my career choice to IT.

[@larsbrinkhoff]

Ye gods weep. I hear Paul Allen wants to add a working KA10 to his Living Computer Museum.

[@MultiShawnt]

This is Alien Tech, forsure. KS KA KI KL ~ SAIL Careful !i