What was the first computer used for IC testing in the 1970s?

[dlgoff]

TL;DR

The first computer that I played with

When I worked for Kantronics, Inc back in the '70s, this was the first computer, the DEC PDP 8, I used to do IC testing after hot & cold temperature soaking, the engineer that could program it with his eyes closed, now works for NASA:

 

[jedishrfu]

A most magical computer! The Altair 8800 and Altair 6800 were pale imitations.

 

[hutchphd]

Summary:: The first computer that I played with

after hot & cold temperature soaking, the engineer that could program it with his eyes closed

That is what stuck after my firsr read...

I remember those but never used one.

 

[dlgoff]

That is what stuck after my firsr read...

I remember those but never used one.

When I was involved with it, we had 50K ICs to test. Needless to say, it took a long time. The guy who works for NASA (I assume he still does; haven't heard from him for years now) could enter data and addresses faster than I could type on a typewriter.

 

[Baluncore]

That appears to be a slow, and low-power PDP-8/S.
If I remember correctly from the 1970s, it had a 1 bit serial CPU, and was used portable, in the back of a truck for seismic surveys.

 

[Borg]

I looked at that and it reminded me how computers used to have a key for security.

 

[Mark44]

The PDP-8 was before my time as far as programming goes, but I do recall reading about it. Nowadays, we tend to think that bytes are and always were 8 bits, and don't realize that in the early days this wasn't always so. The PDP-8 architecture used 12-bit words, so a half-word (or byte) would have been 6 bits.

This was a very simple computer, with only three registers: AC (accumulator), PC (program counter), and L (link register/status flags). Quite different from present day computers.

Wiki article: https://en.wikipedia.org/wiki/PDP-8...es at different points in the operating cycle.

 

[dlgoff]

The PDP-8 was before my time as far as programming goes, but I do recall reading about it. Nowadays, we tend to think that bytes are and always were 8 bits, and don't realize that in the early days this wasn't always so. The PDP-8 architecture used 12-bit words, so a half-word (or byte) would have been 6 bits.

This was a very simple computer, with only three registers: AC (accumulator), PC (program counter), and L (link register/status flags). Quite different from present day computers.

Wiki article: https://en.wikipedia.org/wiki/PDP-8#:~:text=The electronics of a basic PDP-8 CPU has,purposes at different points in the operating cycle.

Yes 12 --bit words seems strange now days. Thank you for the wikipedia link.

 

[jedishrfu]

GE / Honeywell machines in contrast used 36 bit words.

 

[anorlunda]

GE / Honeywell machines in contrast used 36 bit words.

Not all of them. The GEPAC 4010 and 4020 process computers had 24 bit words. I think you're thinking of the GE 625.

24 bits was nice. 8 octal digits.

 

[jedishrfu]

Yes, the GE625. GE635, Honeywell 6000 mainframe machines were the 36-bit ones.

 

[Baluncore]

24 bits was nice. 8 octal digits.

I think you mean 10 octal digits.
Because the first person to say "eight" paid for the drinks.

 

[Mark44]

I think you mean 10 octal digits.

Ha ha!
Sort of akin to "there are 10 kinds of people: those who understand binary, and those who don't."

 

[jtbell]

Yes, the GE625. GE635, Honeywell 6000 mainframe machines were the 36-bit ones.

The DEC PDP-10 also used 36-bit words.

When the U of Michigan physics department retired its PDP-10 and switched to a VAX 11/780 (32 bits) in the late 1970s, I had to write a program (in Fortran!) to convert my research group's data tapes. I had to get down "in the weeds" with the details of integer and floating-point numeric formats on both machines.

 

[Baluncore]

In the early 1970s, the Burroughs large systems were very advanced for their time.
They were stack based, with 51 bit wide virtual memory, but the users only had access to 48 bits of the data, and no access to the address. The OS had access to the 3 tag bits that identified memory use which made the system very secure.
https://en.wikipedia.org/wiki/Burroughs_large_systems

 

[jedishrfu]

with respect to the Burroughs there was a story I heard from our school where an engineer / programmer from Burroughs was taking classes and managed to get his hacker program running.

I think it played Anchors Aweigh on the printer and then started shutting down subsystems one by one freaking out the computer operators on duty. I never found out if the guy graduated or not after a stunt like that. I also suspect that he brought in the program from work and didn’t write it himself. It may be also that this is just an urban legend from my school.

 

[anorlunda]

I think you mean 10 octal digits.
Because the first person to say "eight" paid for the drinks.

Ha ha. Zing. You caught me flat footed.

Some people might be unaware of BCD decimal computers. Notably, the IBM 1401 (business) and IBM 1620 (scientific), both highly successful computers of their time.

The 1620 was said to be, "variable-word-length decimal, as opposed to fixed-word-length pure binary." It used lookup tables for addition, subtraction, and multiplication, and software subroutines for division. My first exposure to computers was on a 1620 in the year 1962. It used punched card input and punched card output. Printing was done by a separate free-standing device. Freshmen students were told to use FORTRAN. Only advanced students got to write machine code.

 

[dlgoff]

(in Fortran!)

I really loved Fortran. When I used it, we had to turn in decks of cards for a program and wait for hours to get the output printout in order to trouble shoot the errors. I also have memories of an EE project where we were given a deck of cards that were all out of order. The assignment was to get them back in the right order.

 

[jtbell]

During our "PDP-10 era", most of our programs were on punched cards. After we switched to the VAX, everybody was "encouraged" away from punching cards on a keypunch machine, and towards keeping source code on the computer and editing it with the newfangled video terminals. ASCII text only, of course. I became something of a whiz with the TECO text editor.

Terminals that could also display line graphics didn't come until later. During my last year or two in grad school (early 1980s), I programmed one of those (still in Fortran!) to graphically display particle tracks through our detector.

 

[jedishrfu]

I think Tektronix had terminals where real line graphics could be done. Basically green lines on a black background, no color.

 

[dlgoff]

I think Tektronix had terminals where real line graphics could be done. Basically green lines on a black background, no color.

Yes. I remember working on those terminals.

 

[jedishrfu]

We used them at GE to draw telco maps as part of the GE US DIalcom network. Each line represented a telpack, a licensing agreement where N simultaneous calls could be made from city to city.

GE would structure calls such that they conformed to the telpacks in use so as not to be overcharged for overuse. As an example, a call from NYC to Boston might exceed the telpack limit at that time so it would be rerouted NYC to Schenectady and then Schenectady to Boston.

At least that was my understanding at the time.

and the dissolution of 1981 Telpack pricing:

https://www.nytimes.com/1981/05/07/business/company-news-telpak-discount-ended-by-bell.html

 

[phyzguy]

I remember we had one of those PDP-8s when I was in school. One time we needed to re-boot it. First you had to enter a set of instructions using those white and orange binary switches on the front. After you had entered maybe 50 combinations (hard to do without making a mistake), this basically installed a simple program that could read a paper tape in the tape reader. Then you read in a long paper tape, which took at least an hour. That basically installed the OS, and you could begin programming it. Those were the days!

 

[Nugatory]

Ha ha!
Sort of akin to "there are 10 kinds of people: those who understand binary, and those who don't."

Mmmmm... No. The 10 kinds of people are those who understand ternary, those who don't, and those who think this a joke about binary.

 

[DaveE]

IMSAI 8080 was my first hands-on computer.

I was in high school. No other devices, no I/O, no storage. So your program was entered into RAM with the front panel switches and disappeared when the power was turned off. You got your output by reading data from RAM locations. No compilers, assemblers, linkers, etc. The UI was access to the data and address buss with LEDS for each. I recall being really, really, good at flipping those switches. OTOH, it was such a PIA that I think my crowning achievement was to write a program to multiply 8-bit numbers.

It was a great introduction to HW, but an awful way to learn programming.

 

[jedishrfu]

No kidding. I had the MITS Altair 6800 and it had the same input method and lack of peripherals. You could buy them but the cost was simply astronomical.

I considered buying a CRT terminal. The cheapest was the Televideo 910 terminal. It was white chars on a black screen and gave you a headache from the CRT hum. I used them at work but preferred the much cooler Televideo 925 but it was too expensive.

The plan was to connect the terminal up to the Altair and use the machines monitor program to enter and run programs. But I decided against it when I heard about the Radio Shack TRS-80 coming to market and felt it was more what I was looking for.

 

[happyhacker]

Summary:: The first computer that I played with

When I worked for Kantronics, Inc back in the '70s, this was the first computer, the DEC PDP 8, I used to do IC testing after hot & cold temperature soaking, the engineer that could program it with his eyes closed, now works for NASA:
View attachment 294005

I worked for BT back in the early seventies with the first broadband (2MB PCM switch) run on a PDP11 doing the call handling. Stored Program Control they called it although that sounds a bit weird now. It wasn't long after that the microprocessors came in with MSI. What processor did the PDP have?

 

[hutchphd]

Don't know where else to advertise this very interesting (to me at least) interview that only a geek or proto-geek would love.

Brian Kernighan interviews Ken Thompson​

 

[Mark44]

When I worked for Kantronics, Inc back in the '70s, this was the first computer, the DEC PDP 8, I used to do IC testing after hot & cold temperature soaking,

The first computer I worked with was in '71 for a programming class I took at my first college. I have no idea who manufactured the computer, as all I did was keypunch a deck of cards (in PL-C, a Fortran-ish language), and submit them between a few Job Control language (JCL) cards. Since programming was something I knew very little about, my programs usually didn't compile, and I would get back about 6 feet of fanfold paper with gibberish covering most of it. Eventually I got the hang of things, and my programs would run and produce reasonable results. I didn't get the programming bug at the time, because it was usually a 24-hour turnaround between turning the card deck in and finding out whether the program ran or not.

My first exposure to computers was on a 1620 in the year 1962. It used punched card input and punched card output. Printing was done by a separate free-standing device. Freshmen students were told to use FORTRAN. Only advanced students got to write machine code.

That was pretty much my experience in two of the first three programming classes I took, in 1971 (PL-C with cards), 1976 (some flavor of BASIC on IBM 370?), and 1980 (Fortran 77 on ?? with cards). The next class I took was an advanced class in Fortran in '82, and the school (Univ. of Wash.) had switched over to VAX minicomputers running VMS. I dropped that class because the instructor admitted that he didn't know Fortran, and he wasn't familiar with VAX VMS.

@anorlunda, your early intro to computers reminds me of the sister of a close friend of mine. She got her PhD in Mathematics, but paid for it by her expertise with computers and programming in the 60s. She (Evi Nemeth) went on to write a very well received book on Unix System Administration. Sadly, both my friend and his sister have passed on - he died of cancer and she was lost at sea somewhere in the Pacific Ocean.

 

[dlgoff]

The first computer I worked with was in '71 for a programming class I took at my first college. I have no idea who manufactured the computer, as all I did was keypunch a deck of cards (in PL-C, a Fortran-ish language), and submit them between a few Job Control language (JCL) cards. Since programming was something I knew very little about, my programs usually didn't compile, and I would get back about 6 feet of fanfold paper with gibberish covering most of it. Eventually I got the hang of things, and my programs would run and produce reasonable results. I didn't get the programming bug at the time, because it was usually a 24-hour turnaround between turning the card deck in and finding out whether the program ran or not.

This is exactly how it was for me as well. I remember sleeping in a waiting area waiting for the output. Then troubleshooting the code, punching more cards and turning the deck back into rerun.