Debugging the 1959 IBM 1401 Computer at the Computer History Museum
ฝัง
- เผยแพร่เมื่อ 28 ธ.ค. 2024
- A little behind the scenes look at what it takes to keep the two giant IBM 1401 "Compusaurs" running at the computer history museum. Today the tape drives decided not to work.
The IBM 1401, introduced in 1959, is one of the first transistorized computers, and certainly the most successful computer of the early 1960's: by 1965, half of the world's computers were 1401's. Only two are known to be still functioning - and here they are.
Compared to my humble restorations, this is major hardware and it could not be done by one person alone - it takes a full team of very dedicated people to keep it going, many of them retired IBM engineers that worked on these machines long ago. So once or twice a week, we get together, fire up the beast, and for a day it's 1960's engineering all over again.
Related Links:
Official video of our IBM 1401: • 1401: The Dawn of a Ne...
Period video introducing the IBM 1401: • IBM 1401 French Presen...
Robert Garner's blog article about the restoration: www.computerhis...
The restoration team's hardcore technical website: ibm-1401.info/#...
Our sponsor for PCBs: www.pcbway.com
Support the team on Patreon: / curiousmarc
Merch on Teespring: teespring.com/...
Learn more on companion site: www.curiousmar...
Contact info: www.youtube.co...
My family visited the museum yesterday and watched the demo of the 1401 and story behind the two machines there. Hearing about all the retired guys that were called to help fix the German machine was my favorite part of the story. I'm sure those guys absolutely love their job. A lot of collective knowledge and wisdom right there too.
I started programming on an IBM 1401 when I was 12 years old in 1968. My Dad was head of the Business Research Division of Southwestern Bell, and he let me play with it on weekends when he was working overtime. I wrote Autocoder and FORTRAN II programs on cards with an 026 card punch. It was a magical time for me. I was taking an electronics class from Cleveland Institute of Electronics at the time. I majored in Electrical Engineering (1973), and also attended Control Data Institute in St. Louis (1975). After that I majored in Physics and Computer Science (1976-1980). In 1981 I became a Field Engineer for Tandem Computers, and then a contract programmer for the next 24 years. I'm retired now, and have fond memories of the IBM old 1401.
4nuk8r I retired in 2004 but right up to that time I was still doing work in Cobol on tandem computer systems. I own an IBM 370 back in the 1970s and we never did get it running. I wish I had kept it I could have donated it to the Computer History Museum. Oh well...
You sir, had an AWESOME childhood!
I envy you sir, it must have been a blast playing around with those big machines
You had an amazing childhood i would but one if i had money
I was a computer operator back in the 70’s at a food distribution center. I ran a 1401 which produced the pull tickets and inventory for the warehouse. The system was not on a raised floor, I was stepping over cables all day long. The 1311 (can’t remember) disk drives were fun to watch, hydraulic arm that would move the heads and would shake the whole unit seeking data. Was quite the work horse and very reliable. That was my first IT job, I eventually ended up working for IBM in 89 for 23 years in systems programming (remember CMS) , great company and people to spend a career with.
Real nostalgia for me. When I joined IBM in the UK in 1970, they were still running internal applications on a 1401 at IBM UK HQ in Chiswick, West London. Even though I was hired as a programmer, all new hires spent some time in operations as part of our induction training. We operated mostly System 360 machines, Models 50 and 65, and we were expressly not let anywhere near the 1401s because of the risk that we might do something to break them. There were CEs on site 24/7 for the 1401s but the 360s were reliable enough to use the same call-in service our customers used. (CE stands for Customer Engineer, IBM's name for hardware repair techs).
At the other end of my 34 year career in IBM, I was leading the team that created the TC/IP load balancer for the Atlanta, Nagano and Sydney Olympic Games web ssites..
Fresh out of the service working at a service bureau, we ran a 1401 with a 1402 card read/punch and a 1403 N1 printer and four 729 M2 tape drives. it was a workhorse of a machine going 24/7 of course we had a parts room with cabinets full of just about any part needed as well as practically live in Customer Engineers. I guess you never forget your first mainframe !!
A 1403 line printer, 1402 card reader/punch, and some 729 Mark II tape drives ... I remember them all well. We had the same peripherals on a 1410 that I operated part-time as an undergraduate in the late sixties. Our 1410 could run in 1401 emulation mode, but we usually ran it in it's native 1410 mode with the PR-155 operating system. The console of the 'ten included a modified Selectric 'golf ball' typewriter. It displayed messages for the operator, and also allowed us to type in the bootstrap loader (starting at address zero). The boot loader was "A word-mark L % B 0 0 0 0 1 2 $ word-mark N." After hundreds of times typing that, the sequence is burned deeply into my 67-year-old brain! Then set the Mode switch to Run, hit Reset, then Start. The L%B000012$ instruction read the bootstrap from tape unit zero, channel 1, binary mode, non-overlapped (i.e. wait for the input to complete), storing machine instructions of the bootstrap itself in memory starting at address 12. Then the PR-155 Resident Monitor would load from the S.O.F. tape on unit zero. (Our PR-155 res-mon occupied the first 11,367 bytes of the 40,000 byte memory - decimal addressing and memory sizing here, not sized in powers of 2 as you might expect). Next, the res-mon loaded the Transitional Monitor (also from the S.O.F.). Transit-mon started reading from the card reader, looking for a MON$$ JOB card (start of a batch "job").
Ah, the power of a fraction of a cheap cell phone, in a package the size of a small drug store! :-)
Thomas Johnson why did tutor want it?
Thomas Johnson A fraction ? Something like one million times slower, and that's being very conservative. Those things were huge and so simple at the same time. I never got to enjoy such machines as I was born at the dawn of the 21st century, but boy do I enjoy learning about them ! :)
a fraction.
1/1000000 is still a fraction.
Amazing... It never fails to astound me just how much the human element was a component of these early "computers". Almost purely mechanical in nature still, operating in similar fashion to today's systems, only taking a microscopic portion of that modern day CPU, and blowing it up to the size of your living room, kitchen, bath, and well... the entire first floor essentially. And having that account for maybe about an atoms worth in a CPU virtual scale sandbox the size of our moon. With the human element flipping switches, dials, choosing to run, stop, assign or take away resources, etc. Things our OS, or the hardware itself does inherently on its own these days... Most don't even remember a day when a motherboards Northbridge encompassed the systems memory controller... But how about when that was "Bob's job". And those memory pathways weren't traces on a motherboard, but cables inside of conduit. Freaking crazy. But everything starts somewhere, and this was always the same thing... Just scaled way way up, or is it down? Scaled.... It was scaled.
As a student computer operator at the University of Texas Computation Center in the mid-1960’s, my job was to operate an IBM 1401 computer system.
The “mainframe” on campus back then was a Control Data Corporation model 1604, an early Seymour Cray machine. The 1401 was used to feed input tapes (typically containing source code, such as Fortran) to the 1604 since the only I/O for the 1604 was magnetic tape, i.e. it had no line printer or punched card reader.
1604 input tapes were created on the 1401 by having the 1401 read user’s punched card decks from its 1402 card reader/punch and then to write the card decks to the 1401’s tape drive(s).
Output tapes from the CDC 1604 (typically Fortran source program output) were read on the 1401 and then either printed on the 1401’s 1403 line printer or punched into blank cards on the 1401’s card reader/punch (the 1402).
Usually my final task of the shift was to match printed and/or punched card output with its corresponding input card deck, which was then all handed back to the person who submitted it.
Students and faculty alike used this process to submit their source card decks for input to the 1604. Most users didn’t realize there was this multi-step process in place for running their programs.
That was the relatively simple then, and compared with the much more complex now, this iPhone 6 I’m using to create this prose is far more capable and faster than many 1401s!
All things considered, it was a fun job and helped put me through college!
Worked on that system at Unilever headquarters in Rotterdam 1962/3.
Many thousands of punched cards passed my hands. I used to carry 3000 at a time in a vertical stack; lots of sorting if you dropped it!
Waarvoor werden de computers gebruikt bij unilever?
i guess one application would have been bookkeeping.
Hopefully they didn't need those punch cards right away of you dropped them!
Watching IBM computers from the 50's and 60's in operation and debugging let me not only gain more respect for the high level of engineering that was needed for them. But it also helps in finding a deeper understanding or appreciation for 1960's sci-fi shows like the original Star Trek serial. Blinking lights, unnamed buttons and switches that only make sense to the initiated, an interior full of exchangable cards and technological marvels that are unseen elsewhere in daily life. Heck, even the Enterprise's serial number "NCC 1701" seems to be a hidden reference to these machines. Star Trek played in 2266, 300 years in the future. 1401 + 300 = 1701.
I could sit here all day long just watching those tape drives go round.
I used to do maintenance at Unbrako in Coventry on a Honeywell machine and one of the tests was to get EVERTHING going. With eight tape drives whirring back and forward it was like an orchestra playing. I wish that I had recorded it..
Absolutely love this... thanks for your work, thanks for posting, thanks to the volunteers. Your museum is on my bucket list!
I had the opportunity of work with IBM 1401 in 1956 at Creole Petroleun Corp in Maracaibo, Venezuela and my first course of IBM Programming was with RPG in 1968. In 1972 I was part of the team that installed ant mantained the PHILIPS DS 714, a huge computer for message handling at offices of SHELL in Caracas, Now at 85 years I'm retired, but my love for computers is intact. Keep up the good job.
Wow, the 1401 i s so beautifully engineered. Truly a thing of beauty
The printers output stats are still freaking awesome! even today!
+WAQWBrentwood They are. I will make a video of it printing one day. Fast and furious (I mean loud)...
nowadays fast Impact printing is done with drums of types instead of chains of types, all letters of the alphabet around a drum, for each position of a line. Whenever a desired letter is at a right position the hammer hits the paper against the type.
i guess the sound is similarly deafening.
Honeywell and others tried, but they could never get the perfect Lines and letters the IBM printers produced.
I worked with these computers at a bank in Toronto back in the 70's. I started out as a lowly 'Burster/Decollator'. As the Decollator, I worked on this big machine (also called a decollator) taking out the carbon paper from the multi-layer print-outs and as a Burster, I used a burster to separate the pages along the perforations after they were decollated. It was a loud (all day bursting paper apart!) and dirty (miles of carbon paper!) job done in a little room off the printer room. I wonder if there are any old bursters or decollators at the computer history museum.
"We need a little bit of cooling" he says, as massive fan noise can be heard in the background
that was funny
Witnaaay well, the 220 volt thing controlled the fan, the black thing mounted on the wall controlled the coolant, probably freon
which is the coolant in AC's
same sound of an AMD CPU Cooler
The 220 volt converter is for the German 1401 since it requires 220 volts to run. The machine he turned on after that is a server room air conditioner
I started working for Univac in January 1970 at NASA in Huntsville, Alabama at Marshall Space Flight Center. We were called field engineers and then customer engineers after what IBM called their maintenance techs. I left Huntsville in 1976 for White Sands Missile Range (WSMR) in New Mexico and continued working on Univac 1108's thru the water cooled 1100/90 until all of them left around 1993. The IBM machines in Huntsville and WSMR were used for accounting and the Univac machines which were 36 bit were the scientific machines. It is hard to believe that in just a few months our first trip to the moon will be 50 years ago. Wish I could go back and do it again.
Oh cool. I remember reading up on the 1401 restoration. I never knew you were partially involved in that. What a fun machine.
I love how these old ibm people come in and volunteer on this beautiful machine
As a retired IBM customer enginer this broubht back memories the 1403 was very noisy because the hammers striking the moving letters made a squick
And 1403's were VERY efficient at moving paper thru them if you had the wrong carriage tape or program was looking for a channel punch that wasn't on it...as programmer back in 80's, I had that occur a few times and still recall running to printer to try and stop it.
"skip to channel 12" ... eventually you'd learn to have a hole punched in every channel somewhere. But not all on the same line, because that would weaken the tape.
The 1403 was such a solid design, it outlived the 1401 by 20 years or so.
+Jim R So true
The carriage was driven by a hydraulic motor. As an IBM CE I worked on a lot of them.
That is why they had insulated covers to minimize the noise. When the cover was closed it was not so bad.
I grew up playing with Dad's tools, tearing apart radios and scattering memore beads and we had work benches made from the guts of these IBM Monsters. We got whole panels of these marvelouse switches and lights and soldered them up and sold them for alarms, those blue panels were made of the most amazingly heavy and sturdy material! And the frames were heavy alloy aluminum.
I hope you can find enough people to keep these beautiful machines running for the next generations. This is knowledge that should never get lost
When I this, I had to show my mother the video and she told me, "When I was working at IBM, we had whole rooms of these things running all the time and people constantly changing tapes and disks out." This makes me want to travel out to see this exhibit so badly, and even volunteer!
yes, todays computers are boring... it was thrilling era, from 1960'' to 1990'' . but after 2000, it started to suck... completly boring, nothing new.
I worked in a modern data center for 20 years. IBM S/390s (or System z as they became known) still used tapes for 'near-line' storage, though they were then cartridges (looked kinda like 8-tracks!). And 'chain' printers (we called them 'line' printers) were still used. Looked a lot like this old one. Loud as hell too!
Absolutely fascinating. Thanks to the whole team for all their work. This sort of thing needs to be preserved just as steam railways have been.
After operating on a mainframe for 10 + years in the 80's and 90's, I can tell you tapes and printers aren't that fun when you use them all the time. When we were on dayshift we had to find upwards of 200 scratch tapes each day for the night shift to use. This was always a chore as out of the list there would always be about 20 that you couldn't find as a tired operator had put them back in the wrong place. Or even worse when the console prompted you to put a certain tape on-line, and you couldn't find it in the tape library, if the tape wasn't laced up and on the drive in about 10 minutes, you would have irate dev staff on the phone wanting to know why their tape had to put up on the drive yet. The printers are amazing, but the thing about the speed they run at is that it makes them prone to wrecks constantly. When the print wrecks it wasn't just a matter of removing the wreck and putting paper back in, you also had to preposition the print, which wasn't always easy, it could take you 10-15 or even 20 minutes of farting around trying to find out where the print had been up to when it wrecked, which you had to do from the oper, which was in another room, although we did get a secondary oper in the print room after a while which we could log in for this purpose.
diecast jam Why in the 80s? Better things must have existed then
_Better things must have existed then_
#1 The tape drives were paid for.
#2 Disk drives and tape robots were ear-bleeding *expensive.*
#3 Laser printers were also ear-bleeding expensive.
#4 132-column green bar fan-fold output is great for a lot of things that laser output sucks at.
#5 "Better" things like Windows and Unix aren't better than the mainframe (and DEC VMS) for production work.
diecast jam Pretty much why, I did not want to become a unix/nt admin in 1995. Too hard a work. And that was only vax systems that I saw during that time. The sys admin at our school, had big giant black spots under his eyes, had a grey skin and smoked like 60 filtherless cigarettes a day. Nope... That was like the posterboy for NOT becomming a sys admin.
That´s amazing! Clearly before I was born - but it´s nice to see those machines running! Thanks for this clip.
4,000 bytes to hold a working program. No operating system, just load your program in on the punch cards, the first 3 being the bootstrap program from where we get 'to boot a computer'. Even so, 4k is a lot of source code to run through the autocoder compiler. After training to program the 1401, I went on to learn 360 assembler, and then PL/1.
Thank you so much to the curators at the Museum , Please keep them going as long as possible
Very cool video! I was amazed by the spare parts available! Should keep it going for a while!
Loved the video. Very interesting! Would love to see more, your commentary was great! Thanks again. I love IBM and vintage computers. Thanks!! Steven
You guys really need to get some 029 card punches. Those are the ones I used back in 1968. I think those were used with the 360 series however.
Here is a story from the Central Valley near Fresno. Brother and I grew up during the Space Race of the 60's. We could name each astronaut, the missions, orbit times, etc. My Godmother left our area for Santa Clara in 1956 and made a life there. She eventually worked at Stanford as a tech sterilizing equipment in one of the med labs. About the time brother was ready to graduate from high school in 1971, she made my dad an offer to get my brother into Stanford.. she could pulled some strings at that time. We were already messing with radio/electronics kits by then. Dad said no.. Brother eventually went to an electronics school (DeVry) in Phoenix.. but quit after a year due to pressure from Dad. When it came time for me to graduate in 74.. same offer.. same answer. When my sister came time to graduate, dad was leaning a bit until something happened in a school in San Jose that you never used to hear of. Someone killed a girl and left her hanging off a door knob on a high school campus. Dad put his foot down and said the cities were too dangerous.. she didn't go. I didn't find out about this until 2010 when my godmother got sick and went into a rest home.. she told me the whole story. She said she had a side business delivering tacos and burritos to some of the offices where these guys were working on this thing called the "computer". She met some nice Hungarian man with a heavy accent. He used to kid her about her Spanish accent. She told him about my brother going to DeVry and he said to send the nice young man down to see him when he graduated. That man was Andy Groves. Her name was Rose Perez. Someone from Stanford may remember her. Her husband was named Marciano.. she called him "Mars".
Brings back a ton of memories . Computer rooms were loud!!!
The 1403 was probably the fastest line printer in the world when it came out, and I'd imagine it would remain the fastest line printer for a few decades. It could feed at 75 inches per second and print 600 lines per minute. The Model 3 could print 1100 lines per minute.
I just had the biggest nerdgasm of my life
I jobbed after secondary school in a big German administration in the late 90ies, and they indeed had one of those still running. Because the government moved to Berlin, they decomissioned it. Hundreds of tapes have been destroyed. I don´t know what happened to the mainframe. But one day I went into the empty room and pressed the button of the cooling device. It was like if a jet-plane started right next to my ears :-D
So cool, a teacher of mine had a similar memory unit as decoration on his desk
I used to know how to use the console controls on one of those. And program in Autocoder. And set up a carriage control tape for the printer. Our IBM service rep. was a frequent visitor to our site.
what a beautiful machine.
I brought those cards from surplus electronic parts stores in the 1960s to salvage the transistors, resistors, resisters and capacitors to build my own circuits. I never saw the computer they came from, now I I have seen it in this video!
II hope there are plenty of detailed videos of these things because something tells me they won't be able to keep them alive forever. Future generations will have to settle for videos.
Yes, the cloud is "forever" 🙂
Out of curiosity, why dont you replace the individual components (capacitors, resistors, transistors, diodes, etc) on the SMS cards instead of trying to find old stock? Is it due to different power levels? Really curious. Love these old computers! Keep 'em in good shape!
I really like how their design has aged. It looks definetely retro, but sleek and clean.
Amazing. It's a thing of beauty. Those tape drives are amazing. I would love to own this monster. I would tell it exactly what it could do with a lifetime supply of chocolate.
I saw this in person last year when I took my younger brothers to the museum, it was great.
Long ago back in the late '80s I once bought an original PDP-11 used. It had 8KW of wire core. And a wirewrapped hex bus backplane. The CPU and core boards were good. But the wirewrap backplane needed some wires replaced. And friend of mine and I spend weeks testing and rewrapping pins on that backplane until we could boot the damn thing. And then all we did with it was set it up as a coffee table. When friends came over we'd input a cylon eye program at the front panel for giggles. It consumed a good kilowatt when running. With a clockspeed of something like 500 kilohertz. It wasn't very useful.
awesome video !!
Were do you get parts for the machine to repair it? I mean it does break down alot so you must need alots of parts.
The boards with transistors could be repaired with modern transistors, as for the other parts. They probably have a stock pile and in the future 3D printing using metal could possibly be used.
+ dantheman We repair only with original parts if at all possible. We have a large stash of parts, new old parts continue to be donated to us (we just received a large shipment of transistors from France - thank you Vincent), some replacement parts come from duplicate machines we have in the warehouse. When we can't find original items, we go through incredible length re-creating missing ones as close as possible to the originals. See for example the recreated brushes for the tape drives in this video: th-cam.com/video/7Lh4CMz_Z6M/w-d-xo.html . Finally we collaborate and exchange parts with other museums: we recently got a re-created print chain for our 1403 from the The Vintage IBM Computing Center in Binghamton. We were floored when we received that one, this is an extremely difficult part to recreate. Fortunately someone in Binghamton still knew about the original manufacturing method used to make them, and was able to resurrect the production process.
That is amazing :)
I am surprised that the computer is ever turned off as when I worked on offshore seismic surveys in the early 1980's turning the data acquisition computers on and off was regarded as an invitation to trouble. I aslso recall flipping toggle switches on the PDP11 navigation computer to re-boot it in a big hurry to avoid having to abort a long survey line and go around again.
It's 15 kW each, so although it's much less power than a single rack of servers at Google or Facebook these days, we would not want to keep it running outside of the demos. Also most of our faults are mechanical, so the less we run it the less wear on the parts.
Hi, I'm trying to make a 3D model of those tape drives, do you think you could take a flash photo of the part where the rollers are and send it to me? I can't seem to get many detailed images.
Really great stuff, just fascinating!
I wish you were keeping a LINC computer running. I would volunteer. And a Linc-8 and PDP-12 to complete most of the history of the LINC.
And a 704, whose architecture suggested the Lisp programming language.
But 1401s are impressive.
Awesome to see these monsters.
Thank you amazing bit of history.
Debugging this is so fun to watch!
I'm going to trade in my IBM Thinkpad for one of those rooms and 20 people lol
Good boy!
A guy I know managed to record some mp3's on a late model tape drive. His gf was puzzled as why did the tape move back and forth while the music kept playing on.
Magic!
How many of these computers would be needed to run a watered down version of Crysis ?
oldtwins probably waaaaayyy too much, lol well above 200, as these probably had a fraction of the calculating power our cell phones have nowadays
oldtwins haha just joking around with ya
Happy to see that a great piece of computing history is still alive. It also helped to win the Cold War!
Holly crap, those memory banks are beautiful.
No idea how on earth they could fabricate that on a production line, or are they hand build&wired?
In that case, must be a nightmare debugging when one made a very small mistake in production
At first they were hand built, then it got progressively automated with incredibly sophisticated and complicated machines. The cores got smaller and smaller, the latest generation of these in the mid seventies you can hardly see them. Still used in the Space Shuttle flight computer.
Amazing stuff! They must have been incredible expensive back in the days.
So much delicate craftsmanship
And they were actually more dense and with a lower failure rate than TTL memory of that time. Back then 4104 and 4116 (4K bit / 16k bit) dynamic chips consumed a lot of juice and ran hot. They tended to drop bits even when working, so you'd need nine for 4K / 16K with one additional for parity. And because tracings, refresh circuitry, and buffer caps all took space to support those RAM chips, an equivalent amount of real core could be packed in the same space. They'd often just blow in a puff of smoke anyway. Which didn't happen with real core. BUT - solid state memory is A LOT faster. And by the late 70s, with 64kbit chips, physical core couldn't come close to its density. Past that... well, the story only got worse for core. Bubble memory suffered a similar fate.
+Paranoid Factoid
Absolutely, early silicon memory did fight the reliability issues you mention. But that was much later (and not TTL). There was just no such thing available in 1959. To put it in perspective, it would take another 11 years before the first 1K memory chip would appear and start to challenge core memory. That was the famous Intel 1103, in groundbreaking MOS technology. The 4k chip you mention (I think you meant the Mostek MK4096) is from 1973. The MK4116 16k chip is from 1976. Core memory lasted for quite a while. It flew astronauts to the moon, and (gasp) was still used in the Space Shuttle flight computer (all 104k of it)!
There is a video on one of the moon mission's where they show a machine wire wrapping a back frame for the computer that did the lunar landings
Hot swappable circuitry, not even 2015 PCs can manage that!
I didn't think an all-transistor computer would need constant debugging, even one that old...
+douro20 Some you expect to fail often: the mechanicals (there is a large and complex amount of it in the peripherals), the relays, the connectors. But also the transistors, diodes and power capacitors do fail on a quite regular basis.
+CuriousMarc Oh yes, the damn germanium transistors. These little guys age quite badly, I can't imagine how hard it is debugging randomly failing transistors on such a computer, especially when you know that there's not always an modern equivalent
That's why there were so many Field Engineers. On call I travelled 33 miles to Courtaulds Coventry to change a transistor on a print position driver card.
Wait, you changed a single transistor on a single card? I thought the whole point of using modular cards was so you could just swap out the card itself and not have to mess with any soldering.
You had to have the card with you to do that. The transistors used were a small varied type used on the range of cards and it was just a matter of a few minutes to do that instead of travelling to get the right card. Still have some in the garage.
Que tipo de cálculos realizaban estos aparatos?? Sólo numéricos ?? Que aplicación tenían ???
More 1401 videos please! I would like to see someone write some code and run the cards for whatever compiler would be on this 1401. PL/1?
+Warp More videos coming. We program in symbolic assembly using IBM original Autocoder. A (very) few of us are able to program directly in machine code and punch it out on directly cards! But common mortals compile on a SIMH emulated machine (goes way faster), we have a nice graphics front end for it, it's like running the real 1401. It's called ROPE, and it's available here: ibm-1401.info/1401SoftwDevel.html. Try it for yourself! Then we generate a punched card deck and run it on the real machine.
+CuriousMarc That is cool. I will try the ROPE thing!
I started at the age of 13 in 1969. Fortran 4 on a 1401 with 16k auxillary RAM. Is this config reroducible? I'm still a programmer BTW. I now do C++ and C# and Python on the side. A slight change.
I love so much this stuff i wish i could have chance to meet this amazing Srs.
How many watts does the whole system consume under load?
1.21 Gigawatts. ;-)
That is so cool. I wish I had this instead of a boring ass modern laptop
in winter it could double as a heating
Wo needs TV or TH-cam when you can watch these beautyful tape machines dancing?!
Now days al this stuff is so tiny as a phone and a lot of faster but this old stuff stays werey cool
There was a card deck that plays “Anchors Away” on the printer! Do you gave such a deck?
What replacement semis are you guys using to replace the originals?
With rare exceptions when we can't find any 1960's parts, we use originals to replace originals. We have a large stash of Germanium IBM transistors.
Are there any modern equivalents? For example, in this video I think I heard you say someone replaced Germanium semis with Silicon, on a card.
So in the 60s my dad worked at a local Bendix division, and they also had a circuit board fabrication plant, locally, and they made cards for IBM Endicott. After seeing you guys servicing the 1401, I realized we had some of those cards in our basement when I was a kid. I had no idea we had parts to an IBM mainframe in our house.
so those tape drives are those big things you always see in old NASA videos (in t he labs, mission control, etc.)
a massive amount of electricity is consumed just to keep those vintage machines powered up + air conditioning
Can you get replacement transistors for this?
+Galfonz Yes. We now have a stash of orginal Germanium transistors. When we find a bad card, we will repair them with original components most of the time, or in extreme cases have some re-made like in my tape drive video.
“Our tape controller is toast, and we are going to try to un-toast it” ☠️☠️ idk why but this is so funny to me
Whats the biggest hurdle to get over when it comes to restoring computers like that?
+Eric Hanchar Many large hurdles really. You've got to find a big wad of cash and mount a rescue operation to move it, build facilities to receive it (it takes 12kW of power!), get all the original documentation and as many spare parts (or in our case spare systems) as you can, recruit a large and dedicated team of specialists to work on different parts of it. And in our case 5 years of work before it was operational again. On this machine with rusty transistors, finding and replacing the 130 failed SMS cards (out of ~2000) must have been a big challenge. Rebuilding the tape drives was another large project, which included building a full tape emulator in hardware. You can learn more about the restoration story in the links below the video.
Thanks for sharing! :)
How did they get it in a basement?
That's amazing.
Crazy. 4k of RAM is about the size of a small 4k TV.
And the TV has 100s of times more memory in one single DDR memory chip.
How do you source 5081 cards these days?
+knoodelhed Difficultly. I understand the very special paper card stock does not exist anymore. We've tried to make new ones but they all tend to jam once in a while. Best is to find stashes of old ones.
Very noisy but beautiful machine with transistoriced cards.
These tapes were very heavy... :(
There was a memory unit that could be added with 12 K of memory. This would allow the unit to use 16 K of memory. It sat outside the 1401 frame.. There was also a RAMAC unit 1405 for disc space. One head on a moving arm that would enter between two disks and go to the correct track. The disks were about 3 feet in diameter and stacked 25 high. If the correct data was printed the printer could play music.
We have both at the CHM! The 1406 memory extension is connected to the 1401, but the giant RAMAC is not. It is demonstrated on its own by a separate demo team.
komputer legend.
My phone is probably 150,000 times more powerful then that entire room.
We went there a few hours there.
I think they had ICs, but they were still pretty new and not very widely in use at this point in time. Everything here is all discrete TTL (transistor to transistor logic). Far and away smaller than the early valve computers, like ENAC and COLOSSUS (and more powerful too), but still not something any old nerd would be running in his basement.
Can it run crysis
I have a hard on for main frames!! They are just hardware porn! Wire and ferrite memory cores are just incredible to behold!
I think I remember when the guy bought it and put it in his basement, he had to cut a hole in to get it in the basement?
You can always tell a '59 by its fantastically excessive tailfins...
My Galaxy S10 phone has considerably more power than that whole room, we are living in 50s SciFi land right now and yet we take it all for granted. I have 2 laptops, ipad pro, iphone 7 and a couple of tablets I can't remember the last time I used
We completely overshot the wildest predictions for computing and electronic gear. But sadly, no flying cars...
8:48 og was saying yikes before it was cool
awesome
4kb memory, my alarm clock probably couldn't even run on that.
Global Warming Skeptic 4K is somewhat standard for microcontrollers. This means your alarm clock has probably 4Kb of RAM !
... If it's a simple one ...
"let's untoast it"
Reminds of Turing machine from Imitation game