Chris EdsallMar 9
Lars Brinkhoff
Happy birthday, PDP-10!
Mar 9, 2023 at 9:20amWeb
Show threadALMar 9, 2023
@larsbrinkhoff I spent my Thurs nights( no classes on Friday) in the basement of the #computerScience building playing on one all by myself. It was much more reliable than the #IBM 370/145 upstairs. When #PDP-10 crashed you just gave it a few minutes and it would come back up and you could keep working, kinds like it sneezed. The IBM one would just lay on the floor hung until an operator came-in in the morning.
Those were the days when computer engineering was fun.
Show threadLars BrinkhoffMar 9, 2023

@mral Nice! When was this?

Do you remember any serial number?

Show threadALMar 9, 2023
@larsbrinkhoff afraid I don't. its was in spring of 1977.
Show threadKarl AuerbachMar 9, 2023
@larsbrinkhoff We had a PDP-6 at Berkeley, I never used it but I was told that it was a precursor to the PDP-10. I have no idea whether that is true.
Show threadBill RickerMar 10, 2023
@karlauerbach @larsbrinkhoff
It is true. Gordon Bell designed the 18/36-bit PDP-6.
They were very rare and nearly dead-ended 36-bit architecture.
The fully 36-bit PDP-10 used the same FlipChip standard as PDP-8, 11, and saved the 36-bit product line.
Show threadBill RickerMar 10, 2023
@karlauerbach @larsbrinkhoff Thus the PDP-10 avoided the maintenance nightmare that was the PDP-6's ALU, that connected the many ALU bit-slice boards with ribbon'cables connecting the edges opposite the backplane, which required removing (& replacing) two ribbons from 18 boards to try swapping one.
and also just ran faster, as higher density allowed higher clock
Show threadMike GarrahanMar 10, 2023
@BRicker @karlauerbach @larsbrinkhoff
The original Colossal Cave Adventure (DEC Fortran 66?) packed groups of five 7-bit characters into 36-bit integers.
Show threadBill RickerMar 11, 2023

@mgarraha @karlauerbach @larsbrinkhoff
DEC Fortran was still like that when I was working in it in 1980, as Fortran77 hadn't been accepted as FIPS yet.
(I recall being annoyed that the unused bit wasn't the sign bit, which would have been convenient to mark last word or not last word, but rather the units bit.)

and DecSystem 10 COBOL packed more, iirc, i forget if it was 6×6-bit characters?
(In theory could fit 7×5-bit old TTY chars, ugh!)

Show threadBill RickerMar 11, 2023
@mgarraha @karlauerbach @larsbrinkhoff
MULTICS and GECOS on GE/Honeywell iron were also 36 bit hardware, as were many of the earliest big iron.
Before the IBM 360, 36-bit was quite popular for precision numerical work.
https://en.m.wikipedia.org/wiki/36-bit_computing
36-bit computing - Wikipedia

Show threadLars BrinkhoffMar 11, 2023
@mgarraha @BRicker @karlauerbach I have that built with the DEC F40 compiler, running under ITS. I found build instructions on YouTube! https://www.youtube.com/watch?v=IXCb_6HpbYc
Incompatible Timesharing System - Adventure

YouTube
Show threadBill RickerMar 11, 2023
@larsbrinkhoff @mgarraha @karlauerbach suddenly I'm wondering at how much work someone took to port that from DEC 36-bit Fortran to IBM 32-bit Fortran G. Did they have to change just the (un)packing code? I might still have a card deck of that somewhere, but I'm not going looking.
Show threadLars BrinkhoffMar 10, 2023

@karlauerbach It's very much true; the PDP-6 and 10 instruction sets are about 99% compatible.

Do you remember if the Berkeley PDP-6 was installed in the basement of Cory Hall?

Show threadKarl AuerbachMar 10, 2023
@larsbrinkhoff Yes, Cory hall - the building with hidden floors half way between the "real" floors. And also in the basement was the Univac SS-90, with magnetic amps, spinning drum memory, vacuum tube printer, and 90 column punch cards with round holes and bi-quinary encodeing.
Show threadM HorstmeierMar 9, 2023
@larsbrinkhoff definitions for a minimum viable product were much looser in those days.
Show threadretropromMar 9, 2023

@larsbrinkhoff An image description would be really nice to have for this post. Here is the main text:

BULLETIN: Power "ON" for PDP-10

Maynard, Mass. 9 March 1967 (FP)

Usually reliable sources reported tonight that the latest machine in the Digital Equipment Corporation line: The PDP-10 was turned on for the first time, with all modules plugged in.

Present at the ceremony were Alan Kotok and Robert Clements, design engineers, and Don Witcraft, and Jim Drew, employees of the company.

No smoke was observed to be emitted by the machine, which made a large whooshing noise. The noise was attributed to the cooling system. Mr. Kotok pressed several console switches and observed lights flashing on the indicator panel. Lights were also observed to flash when a key on the teletype was hit.

Mr. Kotok was quoted as saying "Whoopee, it goes" while Mr. Clements states "Ship It!". Work is scheduled to resume tomorrow.

Show thread🇺🇦 haxadecimal 🚫👑Mar 10, 2023
@larsbrinkhoff
🎶 It's time to celebrate your birthday
It happens every year
We eat a lot of broccoli
And drink a lot of beer
– Weird Al Yankovic
Show threadKernel Bob Mar 10, 2023

@larsbrinkhoff @larsbrinkhoff I always thought DEC 10 Day was October 12th. Turns out it's March 9th.

Kotok used to come around TMRC when I was there in the late 1970s. He always had stories.

Show threadLars BrinkhoffMar 10, 2023
@kbob We also celebrate on December 10th and 20th. The more the merrier, right?
Show threadLars BrinkhoffMar 10, 2023
@kbob How about March 30th as System Concepts day?
Show threadDave W SmithMar 11, 2023
@larsbrinkhoff Maybe on older-timer knows whether it's a coincidence that Bob Clements went by RCC, and that TOPS-10 had an RCC ("range-checked chunk") stop code .
Show threadLars BrinkhoffMar 11, 2023

@dws Thanks, I didn't know about that stop code.

Another similar thing is RPG "rapid program generation", and also Richard P. Gruen.

Show threadMikkoMar 10, 2024

@larsbrinkhoff

"No smoke was observed to be emitted by the machine" 😍

I test my power tool fixes the same way.

Show threadFusselwurmMar 9
@larsbrinkhoff Bob, Alan, Tim, Jim? Unseen University sure has some weird names on their staff roster.
Show threadJyrgen NMar 9
@larsbrinkhoff By the way, I had the opportunity to (virtually) meet Robert Clements in some other social network place where he participated for a number of years. Totally nice and interesting guy. He stopped posting some years ago. I hope he is well.
Show threadDavid, a Bostonian in Tokyo.Mar 9

@larsbrinkhoff

Do you know what the relative speeds of the PDP-6, KA-10, and KL-10 were?

The MIT AI Lab's PDP-6 was quite stable at about 1/3 of a MIPS. Maclisp was fine on a KA-10 and really sweet on a KL-10 ('73-'76 period).

Later ('82-'86) I was using Yale's Scheme implementation ("T") on 68010 and 68020 workstations. On the 68010 workstations it was horrifically slow, but quite fine on the 68020. The internet claims the 68010 was a "1 MIPS machine", but my actual experience differs...

Show threadLars BrinkhoffMar 9
@djl It should be possible to look this up somewhere. For one thing, speed of main memory is important. But I'm guessing ball park 6 to KA10, 1.5-2x. KA10 to KI10 2x. KI10 to KL10 3x.
Show threadRich AldersonMar 9

@larsbrinkhoff @djl
The KL-10 is roughly a 1MIPS machine; the microcode engine is about 5MIPS.

The KI-10 is about half as fast, and is of course a hardware implementation with no microcode.

I think, but am not certain, that the KA-10 is about 300KIPS, again in hardware.

I have no figures for the PDP-6 other than the clock speed, which was 2 microseconds per cycle (quoted from the magazine advertisement at
https://www.panix.com/~alderson/PDP-6_advert.jpg
).

Show threadDavid, a Bostonian in Tokyo.Mar 9

@alderson @larsbrinkhoff

Thanks. I guess the MIT AI Lab PDP-6 may have been hacked up to run faster, or maybe the 1/3 MIPS I remember is wrong. Or was the microcode speed.

Another problem here is that for running Lisp, the PDP-6/10 instruction set is seriously optimal. That the 68010 is claimed to be a 1 MIPS thingy, but in real life was horrific for Lisp, was that it was "1 MIPS" for 16-bit arithmetic and a tiny fraction of that for slinging around pairs of 32-bit pointers...

Show threadRich AldersonMar 11
@djl @larsbrinkhoff The PDP-6 was not a microcoded device. It was an asynchronous hardware implementation, using germanium transistors on big honking boatds.
Show threadLars BrinkhoffMar 11
@alderson @djl KA10 was also asynchronous. The KI10 was a synchronous design. The KL10 was microcoded, largely inspired by SAIL's unfinished Super Foonly project.
Show threadRich AldersonMar 12
@larsbrinkhoff @djl “inspired by” is an odd way of saying “purchased” 😁
Show threadDavid, a Bostonian in Tokyo.Mar 9

@larsbrinkhoff

Unrelated 36-bit computer trivia.

Today I learned that the IBM 701 was a 36-bit, 16,000 addition operations per second, vacuum-tube computer, and was released the year I was born.

The all-relay Harvard Mark II (5 years earlier) could do about (drum roll) 24 operations per second.

Thus supporting my point that vacuum tube computers were about 1000 times faster than relay computers. We had to wait a long time after the 701 (1952) to get another factor of 1000.

Show threadploumMar 9

@larsbrinkhoff @cstross : today, I learned that I share my birthday with the PDP-10!

(already knew about Youri Gagarine and about my son but that one is new)

Show threadMason Loring BlissMar 9
@larsbrinkhoff Some days my eyes don't work well enough for me to read that stuff so I'm grateful for alt text. There are folks where that's every day. If you could add the text I'd be grateful. I'm not old enough to have played with a PDP-10 but I know their place in our cosmology!
Show threadLars BrinkhoffMar 9
@mason Good idea. I have updated the post.
Show threadMason Loring BlissMar 9
@larsbrinkhoff That's awesome. Thank you. (And thank you for adding the alt text!)
Show threadGemma ⭐️🔰🇺🇸 🇵🇭 🎐Mar 9
@larsbrinkhoff was this written by Douglas Adams?
Show threadLars BrinkhoffMar 9
@gcvsa I can make you a PDP-10 program that computes the meaning of life, universe, and everything.
Show threadLars BrinkhoffMar 10
@gcvsa
Show threadChris EdsallMar 9
@larsbrinkhoff I'm (hopefully) going to use the line "No smoke was observed to be emitted by the machine, which made a large whooshing noise" when we power on our next supercomputer.
Show threadRich AldersonMar 9
@larsbrinkhoff I noticed that the original post was 3 years ago, including a leap year
(365x3 + 1)