Computer History: IBM System/360 Mainframe 1964 ORIGINAL ANNOUNCEMENT, Transistors, Data Processing

[@CMDRScotty]

As someone born in 85 I find computer technology from before I was born so fascinating. Thank you for the uploads.

[@ComputerHistoryArchivesProject]

Hi Steven, you are very welcome! ~ Victor

[@knightwatchman]

It was an amazing time. I began my IT career at 20 years old operating a 360/40. I left a college liberal arts program to take that job. I then go into systems work, meaning I installed operating systems, subsystems, applied maintenance and tuned systems for performance. All this work to ensure the applications programmers could code and test applications and operators could run them. I did this for 14 years and was hired by IBM in 1987. I retired from IBM with 32 years of service. All totaled I had a 46-year career in IT. All because of mainframes.

[@garyclouse7234]

Ever held a magnetic core memory array in your hands? I have! Obviously I am an old person. Punch cards, teletypes, tape drives etc. Now I have a smart phone I could never have conceived in those days!

[@skuula]

Me too, but I'm sure happy that I didn't have to work with the early and not very good negative number representation in those computers.

[@wlellis001]

The introduction of this machine was the reason I got into IBM in 1966. The machines had to be "completed" in the field, in order to meet their shipping commitments. I and 30,000 technicians like me were hired very quickly to install the Engineering Changes necessary to make them actually work!

[@RetroJack]

These days that would be called a day one patch! 🤣

[@joshr277]

Wow! Thanks for sharing - it’s amazing how these pioneering innovation changed everything

[@Lazarus0357]

Thanks for loading this video! I started my career in computers with a 1401, then when I moved to a 360 it was like moving into the realm of magic! Thanks again, you remind me of my mis-spent youth. Regards

[@rty1955]

I began on a 407 -> 1401 -> 360/30 with only 48k of memory. The things I could do with such little memory! I ran BPS & DOS

[@jordanzish]

If you haven't seen it already, check out Curious Marc's channel. He works at a computer history museum and has a series on the restoration of a 1401 system. I think you'd definitely get a kick out of the final video where they have it run a FORTRAN program off punch cards.

[@royf.9034]

"....up to eight million characters of storage. Eight! Million!" Reminds me of when I got my first hard disk: it had a whopping 10MB of capacity. No more swapping floppies! "Ten megabytes!" I thought, "I'll never fill THIS up. I'm fixed for life!" Now a single file would be too big for it, and a 8GB of RAM is only until I get more.

[@EnergeticWaves]

Nothing short of amazing. A single video file can easily be 8gb!

[@peterr1088]

Roy F. "Ill never fill this up" i cant find an app smaller than 10MB

[@cyrex686]

I used to know what every single file was for on my 10MB hard drive :/

[@snow6937]

@@peterr1088 just for example - sumatra pdf (and more types of files) with 60 langs and hotkeys - unpacked disk space - 'bout 7 mb

[@scottcupp8129]

It really is astounding how quickly technology has advanced. I, too remember my first hard disk. It was 10mb as well. And my thinking was the same: I will NEVER fill the drive to capacity :) Now I have a machine that I built for music production. It has 3 internal hard drives and I have 4 external drives for storing my sample libraries. I have a total combined storage of 52 TB. I have 64 GB of DDR4 RAM and dual i7 processors. It is a beast!

[@reverenddick8562]

Found out my 74 yrs old grandpa was one of the guys programing these early machines. What a trip.

[@ComputerHistoryArchivesProject]

Black Sheep, sounds pretty awesome. Very cool. ~ Victor, CHAP

[@reverenddick8562]

@@ComputerHistoryArchivesProject not as cool as him knowing about the hidden Code in the IBM 5100. 😉

[@metafis2490]

My first job after I graduated in 1970 was as a Trainee programmer for Barclays bank (London). 360/65.....main memory was 500k in rows of cabinets...:)

[@johnlister]

1978 NatWest at Goodmans Fields. 370/178 & 3033s with 4MB each. And my phone...! Those were the days!

[@neo214]

Wow, was it anything like working with computers today?

[@allanrichardson1468]

@@neo214 You didn’t “talk” to computers, except on remote terminals using time sharing systems or business remote applications. You wrote the computer “letters:” programs, data files, job scheduling and control files (really programs in their own right), etc. Originally you typed these “letters” on 80-character punch cards (later, by creating or modifying files on disk devices using a time sharing system and terminal), submitted them through a card reader (usually to an operator) or with a “submit” command on a time sharing system, then waited for the job to run and its output to print (later, you could read a copy of the printed output on a CRT screen before, or instead of, printing it on paper; or print it on a remote printer if you could wait that long). Some of these procedures are similar to development on personal systems today, except the processing and storage took place on a shared large machine, not inside the device on your desk!

[@Drew.DrivesYT]

I currently work as an operator on the z/OS system, this is a cool history lesson!

[@ComputerHistoryArchivesProject]

Cool! z/OS is quite an advance over the early giant machines. What release?

[@zorinlynx]

This is so interesting! Solid Logic Tech sounds like it was a bridge between discrete components and integrated circuits; they couldn't quite make true ICs yet but you could tell it was going in that direction. Thanks for posting!

[@ComputerHistoryArchivesProject]

Great observation! Thanks for your feedback!

[@michaelmoorrees3585]

Those are early hybrid circuits. Think surface mount components on a ceramic substrate, instead of a PCB. I saw my first MLCC "chip capacitors" on 60s vintage hybrid modules. Resistors are sometimes painted on with resistive inks. The 360 came out in the 60s. Early logic ICs were available, but IBM was very "vertical", and used their own "in house" semiconductors, almost exclusively. Also those early ICs were pretty pricey, too, and used mostly on government aerospace projects, with deep pockets.

[@LMacNeill]

Yeah, the first CPU-on-a-chip wouldn’t come out until Intel created the 4004 in 1969 - and then it was only a 4-bit processor. Far too small for any real work. It’d be the mid-to-late ‘70s before CPUs on chips would be popular in mainframe computers like this one. The big companies continued to make their own CPUs similar to the one in this System/360 - the CPU chips were relegated to small, desktop-only machines for a long time.

[@bertoid]

That's correct. And System/370, released about 6 years later (in 1970), used very similar looking packaging, but the modules in S/370 had true integrated circuits on the ceramic substrate. The only obvious differences were the module packages were usually larger, and circuit cards were about twice the size. My page on IBM circuit packaging evolution: members.optusnet.com.au/intaretro/Packages

[@rty1955]

bertoid the 4381 was designed so that each module had the same thermo design and was cooled with two automotive fans. The therory being that you can cool a house with just one fan if all the rooms had the same design and temp. It was call "impingent cooling" IBM was always looking to improve cooling. They wanted to get away from big room that needed tons of air conditioning

[@ImNotADeeJay]

Those who live in the Bay Area and want to see one of these in person, I encourage you to visit the Computer History Museum, that happens to be very close to Google HQ

[@taylorp535]

Yes!! This place is super cool!!!! They have all kinds of vintage computers and all this history, totally worth a visit ❤️❤️

[@tecteam]

The ram looks organic almost futuristic, like a woven blanket that moves - so cool

[@collinriley4976]

Brings back a lot of memories (no pun intended). Got out of the Navy in 1966, and after a time got a job as a computer operator with a bank. They were using an IBM 2010 (running 360 emulation), a 1401 for data capture, a 360/30, and a 360/50. All great fun with blown programs, unbelievable backlogs, jamming OCR check readers, rubber band fights with the key entry people, and whisky bottles found under the raised floors--oh, and absolutely no security. Anyone could walk off the street and right onto the computer platforms. The bank ran three shifts a day, probably 20 operators, updating customer files during the 12a to 8a shift; all it took was an error on one of the master file tapes to set us back several hours while we reran the previous night's update. Once management decided to purchase less expensive Memorex disk drives, and that equipment almost crashed them out of business. Finally got out of there, after 10 years (and several system changes) to go into tech writing.

[@ComputerHistoryArchivesProject]

Great memories. "and whisky bottles found under the raised floors--" that one made me laugh. Thanks for sharing!

[@madezra64]

Lol at the whiskey bottles. I think a lot of people underestimate substance abuse among hardware and software engineers as a whole. NOTHING beats staying up for days programming all stimmed out... ANYWAYS that's my modern whiskey bottle hidden away moment :P (Used to stash meth under my office desk drawer for those long ass binges where sleep was impossible to get cause of all the constant calls I had to support...)

[@rty1955]

Haha I worked at the nations largest bank and even though we had security guards and our data center was locked, the combination was 1234!! One data center I worked in was 200' x 200' with (2) IBM 370s, a few series 1s, and lots of other vendors. The ops always had beer under the floor! They used to shut off the lights at night and have ring toss fights (the write rings from tapes) in the morning I would come in (I was the systems programmer) and see the rings all over the place!

[@samsungtelevision695]

@@madezra64 you’re not wrong. Usually these days it’s adderall (so I hear) but they can fill the same niche for sure

[@mikeklaene4359]

I learned programming on a 360/30 with 32K of memory. It was in 360 assembler with DOS/360. Later went to a company with a 360/65 with 512K running under OS/MFT - still mostly assembler with some PL/1. Fun days!

[@ComputerHistoryArchivesProject]

+mike klaene Very cool! I consider anyone who could program in assembler, a highly talented individual. "Shift RIght, Double Algebraic" - those were the days!

[@usnay12345]

+mike klaene Wow, I didn't realize I was living so high on the hog -- we had two 360/30's with 64K each. DOS/360. COBOL shop. Started teaching myself 360 assembler ... finished that on 370/138 with 512K. Got to spend 6 years FT assembler work, early 90's. Most fun language I ever knew! Wanted to try PL/I, but we never had a compiler. If I can find all the pieces, maybe do something with Hercules ...

[@mikeklaene4359]

usnay12345 It was at a dept store in Cincinnati where I got my start programming. No degree at the time - pure OJT via some good mentors and the green and white IBM 'Programmed Instruction' books. The HQ for our parent company was across the street from our store and they had a 64K 360/30 which we had to use to do the COBOL stuff. These days there are a lot of coders but very few programmers, IMHO.

[@ComputerHistoryArchivesProject]

+mike klaene Hi Mike, Assembler on the 360 must have been quite interesting. I did a little Assembler on the IBM 4341. "SRDA, BAL, etc." Brings back memories. Thanks for watching!

[@royf.9034]

I'll bet those were the days! You were there at the invention of modern computing.

[@sfperalta]

I started studying "computer science" in 1977, though there was no curriculum called that at the time. Most folks who entered computer programming came from other disciplines - math, chemistry, business, etc - where they were exposed to computers and coding as part of their work and got hooked on coding so switched to doing that full time. I started with electronic circuit design and ended up switching to programming full time 5 years later. The IBM 360 was still a strong influence, through supplanted by the System/370 by then, which was operationally about the same as the 360. You never actually operated the S/360, but interacted through 3270-style terminals which could be in the same building or miles away. Coding was a trial and error process often taking several days to write, submit, wait, get result, fix bugs, then repeat the process all over again. Still, it was easy to get hooked. When I retired from software 7 years ago, most programmers were actual computer scientists (BS, MS or PhD) and the average personal computer was many orders of magnitude faster than those original 360/370 systems. I can't imaging what computing will look like in another 40-50 years!

[@ComputerHistoryArchivesProject]

Hi @sfperalta, thank you very much for your fascinating story of your experiences with computers back to the 1970's. It sounds like your branching out into electronic circuit design and then to programming was a great way to get a great overall appreciation for systems that were evolving, and the software behind them. Thank you for sharing this look back and for your insights into some of the changes you have seen in the industry. Quite interesting! ~ Victor, CHAP

[@threadracer6076]

"One system to meet every application need." "One ring to rule them all!"

[@ComputerHistoryArchivesProject]

+Thread Racer LOL! Great quote, and so applicable! : )

[@batlin]

Or a few hundred thousand "rings", in the core storage wiring... :)

[@michaelmoorrees3585]

In the early 80s, we had a project that used 16K bytes of core, on a single Multibus card. Multibus was common board form factor from the mid 70s into the early 80s. Core is non-volatile, which was its only useful feature, at that time. In a year, or two, EEROMs (not EPROMs) became common, and killed that last use. Seeburg jukeboxes used core memory, for a time. Its a perfect app for core. In a jukebox, the entries in the play queue are erased, when that song is actually played. With core, the process of reading erases, the read location (destructive read), so that step, is automatic.

[@oldguy7402]

Or IBMs token ring network...

[@AliasUndercover]

That's the CORE memory.

[@sybillestahl8646]

Oh my goodness! My highschool science club went to Waterloo University in, probably, 1968, did a short introduction to Fortran programming, and marveled at the IBM 360 75 in its sunken glass-walled room. So many changes.....

[@michaelc1148]

The closed 'the pit' at U(W) back in '99. (I was an undergrad there at the time). It's now a bunch of lecture rooms.

[@supercompooper]

I have a tile from the Red Room 🙂

[@sski]

This brings back memories. During the middle to late 80s I worked for Spencer Gifts, Inc. at their 5150 building in Linwood, NJ as a laser printer operator. Spencer would mail out catalogs with personalized covers which we would print from magnetic tapes running from a mainframe. It was all so "cutting edge" then working with an industrial computer system. Seems archaic now when one box the size of a toaster can do the same thing and more, faster. But I wouldn't trade the experience for anything. I can still remember the smell of walking into the printing center for the beginning of each 12 hour shift. Ozone and ink. Ahhh.

[@ComputerHistoryArchivesProject]

Thanks for the cool comment. Wow, a 12 hour shift. I hope they let you bring plenty of snacks onto the raised floor. Nothing helps pass the time like Cheetos, Pringles and coffee...

[@ComputerHistoryArchivesProject]

Most large data centers of that time, and up to recent times, had a "raised floor" with 10 to 15 inches of space under the floor tiles to hide all the power cables and data transfer cables. Alternatively, they use overhead cable raceways to hide all the cables today. Cable and wire management is a true art in itself! Good comment. Thanks!

[@rty1955]

the reason you used raised floors was not to "hide" the cables. it was because the air conditioning was forced down I to the floors and up through the equipment. no need to freeze the operator by making the entire room too cold. more efficient to have the a/c flow through the equipment then out into the room where it would be comfortable for people operating it.

[@knidhi8993]

Yes, it was known as false flooring like these days false ceiling.In south east Asian nations, false flooring is a normal way of lihome building to keep the insects and smaller animals not coming to usable surface.

[@typograf62]

I have often considered that for my home computer installation, but the door opens inwards and makes the idea impractical. Perhaps I should run cables down from the ceiling? They often had a special suction cup device for lifting the tiles.

[@ComputerHistoryArchivesProject]

I tend to agree. Under floor electrical power cables and data cables kept them out of the way. Cool air flowing upward into the center of the machines was often more efficient than blowing it around the room. Have also seen newer data centers that have data cables running along the top of the machines, just below the ceiling, that gives easier access to cable changes and fixes, than working under the floor in tight spaces. Progress!

[@shmehfleh3115]

The newer datacenters I've worked in tend to run power in the floors and data in overhead raceways. Makes sense, I guess.

[@allanrichardson1468]

Although I was a S/360 and S/370 programmer for decades, I just now noticed the symbolism of the original System/360 logo! Prior computers grouped bits in triplets and multiples of 3: octal (base 8) digits to display groups of 3 bits; 6-bit text character codes; 6 characters per 36-bit binary “word,” etc. Just like the 360 degrees of a circle (thanks, ancient Sumeria!). But IBM showed by its compass rose logo that those 360 degrees of data representation would hereafter be grouped by 4s, 8s, 16s and other powers of TWO: 4 bits displayed as a hex (hexadecimal; base 16) digit, using the letters A through F as the extra six digits; 8-bit characters, also called “bytes” (an IBM-invented word), divided into 4-bit “nybbles” (also IBM-invented); 2-byte (16-bit) “half words;” 4-byte (32-bit) “full words;” and 8-byte (64-bit) “double words.” And the four cardinal points and four secondary points in the logo, eight in all to complete the circle, represents this shift in thinking!

[@ComputerHistoryArchivesProject]

Hi Allan, that is a fascinating analysis. Never saw it broken down like that before. Thanks! ~ Victor, at CHAP

[@waltz49]

I started my career in supporting processor development for the S/370 and ended it working on emulated System z mainframes that could support lots of users on a regular PC. I find it really interesting to see how people compare the 60s and 70s tech and environment to today's. In those days, when processor innovations like instruction pipelining and memory hierarchies were new, it was a major achievement to get a 10% performance increase or optimize an algorithm to free up cycles or memory. Now we take for granted, and often can't even notice a doubling of performance every few months and memory constraints are a distant memory (hmmm). Those days were fun because the possibilities were endless and you had to know how things worked. Now programming is about libraries and frameworks and fashions and fads and hardware constraints don't even exist. Bleccchh. Great vid. Thanks (oh, and yes... Get off my lawn!)

[@ComputerHistoryArchivesProject]

Hi waltz49, thanks for your comments and thoughts. I agree with your statement on how previous memory/hardware constraints don't even exist today. It is also fascinating to see how the concept of what a "computer professional" is today, vs. what the term meant back in the '60's and 70's. I think back then, it seemed one really had to know math, engineering and related disciplines just to understand and follow the early textbooks. - PS, I promise to stay off your lawn. :)

[@Bhakti-rider]

@@ComputerHistoryArchivesProject Nah, mainly logic; a gift for mathematics of course helped a lot.

[@Bhakti-rider]

A systems programmer I knew at Stanford Hospital in the early 70s made the comment about some piece of software (it might have been TSO, which I hadn't met yet; we used WYLBUR), "TSO [or whatever it was] may be big - but it's slow." That got a tremendous laugh, from me and everyone else who heard it.

[@saskiavanhoutert3190]

The history of IBM computers is relevant for further prospective, i just love it.

[@stanscocom]

Reminds me of when I bought a 20 Megabyte Hard Disk for my IBM PCjr (around 1986). It was $700 and was the size of a toaster. I remember thinking "I'll never use all of this space!". I fly drones now and less than 2 seconds of 4K video from my drone is more than 20MB.

[@mrs7195]

It is kind of uncanny to realize that despite how much storage space you have, you can always find ways to fill it up somehow. I'm just a regular computer user with a 1 TB hard drive, but I have less than half of it free anymore and I have no idea how that happened.

[@ComputerHistoryArchivesProject]

PLEASE JOIN US in Preserving Computer History with a small contribution to our channel. www.paypal.com/donate/?hosted_button_id=LCNS584PPN28E We get no revenue from Ads, KZbin takes all that revenue. Your support is what keeps us going. Your contribution greatly helps us continue to bring you educational, historical, vintage computing topics. Thank you! ~ Computer History Archives Project

[@beagle7622]

I love this stuff. My dad’s career grew with computers from the early 1950’s. He ended up in charge of a very large department of around 300 people . It was a fascinating job that he loved. The IBM computer stuff I saw as a kid, the 360 with flashing lights fascinated me. They had a couple of 30’s a 40 & a 50 before upgrading to 370’s.

[@ComputerHistoryArchivesProject]

Hi Beagle76, thank you very much for your feedback! We hope you will explore some of our other vintage computer videos as well. Keep well! ~ C. Hunter, CHAP

[@beagle7622]

@@ComputerHistoryArchivesProject I am . Dad would have loved this . I remember the Ramac very clearly, he talked about it a lot when I was in school. I think there is a computer museum in Melbourne Australia. The last computer I saw with him were 2 Armdahl’s. A smaller one controlling the big. He talked about doubling the memory from 8 to 16 Mb on an IBM I think,yes megabyte & how much it cost. He started off with ICL’s I believe.

[@EnergeticWaves]

Gotta love that 60s style music.

[@thepain2222]

2:17 "A single system featuring unprecedented performance range. New micro electronic circuitry. Over 8 million characters of core storage!" LOL Great stuff and fascinating. Imagine what the IBM of that time would say about the capacity of a single micro XCSD card. Many orders of magnitude more storage in the area about the size of a pinky finger nail.

[@ComputerHistoryArchivesProject]

Ron John63, yes, "Channel Capacity." Good point! ~Victor, CHAP

[@Musketeer009]

Sorry, IBM, you did not herald the opening of the 'Computer Age'. Colossus did that in Bletchley Park UK to help break enemy codes during WW2.

[@TechnikMeister2]

So True. Ken Olsen and Digital was doing this stuff in the 1950s. IBM 360 were just simple database machines. They were used for bank account records and inventory management. Very simple but high volume stuff. They were strictly batch processing. It was not complex calculations and engineering solutions like Digital who also had realtime processing. You could update the operating system of a PDP 11 without shutting it down. It was MIT with Digital and Harvard with IBM. That tells you the emphasis.

[@fredblonder7850]

I used one of these installed in an old dairy-barn. It had a bare dirt floor, with the cables run across it. The CPU and console was in the center, and the card-reader, card-punch and printer were each in one of the cow-stalls. The keypunch machines wire in the hayloft, so you had to climb a rickety wooden ladder. This was at Catonsville Community College (now CCBC) circa 1974.

[@ComputerHistoryArchivesProject]

Hi Fred, that is one of the most interesting locations I ever heard of for a 360 install! Thanks! ~ Victor, CHAP

[@ImNotADeeJay]

7:52 ferrite cores memory... it's been a long way

[@WarrenPostma]

this is great! thanks for posting.

[@ComputerHistoryArchivesProject]

+Warren Postma Thanks very much for your kind words. Glad you enjoyed it.

[@warp9988]

+Computer History Archives I would love to help make videos like this. I am keeping my eyes open for folks who I could interview who worked in computers long before I was born in 1970.

[@benghoor4437]

1:32 Like Cortana promo))) 8:22 This type of memory is not obsolete. This memory is resistant to electromagnetic impulses, it is used in intercontinental ballistic missiles, which allows them to take off in the face of an enemy nuclear attack.

[@ellsworthm.toohey7657]

all these concepts still used today !!!

[@allanegleston4931]

when i was taking computer courses in college, the porp brought out one of those disk packs to show us . they were massive . now the disks are smaller than ur thub.

[@MarkJones-mm3br]

I'm convinced! Where can I buy one? ;-)

[@y11971alex]

+Mark Jones I think they're all out of commission now.

[@MarkJones-mm3br]

+y11971alex It's OK, I was being ironic :)

[@y11971alex]

It's the fact that IBM mostly leased these machines, which means they collected these computers at the end of their leases and threw them away or did something else to them (perhaps someone from IBM can tell us what happened to off-lease computers). There are a few people who collects IBM System/3 (1969) though, and their collections are available on the Internet.

[@MarkJones-mm3br]

It's just amazing and kind of amusing to me to hear them breathlessly extolling the greatness of features that, due to the incredible speed of the advancement of computing, seem so paltry now. Also, the idea of someone today falling for the marketing just because it was delivered so convincingly just tickled my funny bone - a little comedic scenario that played out in my head as I watched the video. :-D

[@y11971alex]

Mark Jones Mainframe programs of the age tend to be very memory-conscious, which is to say that they achieve the same performance with less memory usage. While raw performance definitely falls short of a consumer-grade computer today, you may still be surprised that it isn't so puny after all compared to today's computers. I understand that these computers can go without reboot for years on end, which puts to shame the most stable Microsoft-based PCs and servers today.

[@dd_ranchtexas4501]

Lots of memories for a retired chip designer. Oh and those 1.5 inch wafers with only a hand full of good die. We still had a LOT to learn! Yep the Bad Old, Good Old days!

[@ComputerHistoryArchivesProject]

Hi DD_Ranch Texas, glad this helped bring back some memories. Thanks for the feedback! ~ Mark

[@xapie128]

I worked on an XDS computer in the 1970s that had ferrite core memory. We can all laugh at the amount of memory that seemed "large" back then, but core memory was extremely FAST.

[@w6wdh]

Core memory also preserved data when powered off. I had a 16 bit minicomputer with core memory in 1972. When you turned it off, it wrote its TTL registers to core and shut down. When you turned it back on, it restored the registers from core and continued where it left off.

[@FiscalRangersFlorida]

I was a computer operator for the City of LA in 1971 on a 360 and then a 370 after I returned from Vietnam as a computer operator on the IBM 1800 plotter in Saigon. The 360 was a great machine and I wish I had an original 360 or 370 Console. When I was at the City, the computer operators were out of date and were running only one partition at a time to spread out work load. The 360 had five partitions and so I worked the midnight shift (I was in college during the days) and I once ran the entire next day's production schedule overnight using the five partitions. I was lectured by the union focused employees NOT to do that again. So the city used about 4-5 times the operators and machines (they had 3 360's) that they really needed. That is how your government works.

[@ComputerHistoryArchivesProject]

Hi @FiscalRangersFlorida, That is a great story! Thank you for sharing your experience with the 360 and 370 systems. They were as you know, the powerhouse systems of their day. Massive technical marvels (for the time) and even color-coordinated to meet customer's requests. I wonder how many IBM and other computers the U.S. left in VietNam when we finally pulled out. Probably all of the mainframe stuff got left behind (?). Not to mention our allies and supporters. ~ Your comments about how your making efficient use of the system's partitions to speed up processing time was not well received by management. I understand your position. Good point. Thanks again for sharing this great bit of history! ~ VK

[@altamiradorable]

Used to program on the 360 in assembler language ! Boy was that machine a real beast of a workhorse !

[@rty1955]

I still code in assembly! I wrote complete accounting packages (A/P, A/R, G/L, PAYROLL, PURCHASING) on a 360/30 in 32k!! I would love to see any of todays programmers do that

[@ComputerHistoryArchivesProject]

Roy Yung, Assembly Language is hot stuff. Pretty impressive. VK, at CHAP

[@rty1955]

@@ComputerHistoryArchivesProject yes. To me, its Assembly, then everything else

[@techresearch5901]

I remember seeing this computer on one of the Mad Men episodes. Season 7, episode 4.

[@bjornwegener3233]

"Online" back then meant "on the processing line".... funny how we see it differently now.

[@rabidbigdog]

IBM (with the 360 family) showed that backward compatibility eventually beats out innovation which eventually also proved true with personally computing.

[@HillbillyRednecking]

8 million bits = 1 megabyte... Wow that's a lot of memory!!

[@Bhakti-rider]

1 MB is 8 million bytes, not bits...

[@Bhakti-rider]

Oops - 8 MB, I mean

[@Geert365]

The robotic arms (3:09) seems gentler, from what i have seen in more modern times, these give me the creeps with the how fast, and at the end abruptly slowing down motions, but are more efficient of course.

[@JimGardner]

Really looking forward to all of this coming true. Wait...

[@jourwalis-8875]

Fantastic achievement!

[@ComputerHistoryArchivesProject]

Yes, quite an achievement. Thanks!

[@jackilynpyzocha662]

I took a computer course(intro) at a local technical high school, 1983, that had a 360, in blue. There was a plastic cover over the dot matrix printer, it was loud!

[@bloqk16]

In this video @10:10 the narrator spoke of "power of growth" and went into the subject of how this IBM mainframe system can grow with adding hardware to the system. I wonder if any companies back then had issues with electrical power shortages when additional IBM equipment was added without an upgrade with the incoming electrical supply. I worked for National Advanced Systems (a National Semiconductor subsidiary) back in the early 1980s that had a computer room where engineers kept adding mainframe computer equipment similar to what's in this video. It resulted with an overload to the electrical transformer to the computer room, resulting with it overheating and nearly catching fire. As a stopgap measure to minimize the overheating hazard, an HVAC contractor was called in to install an air duct with the chilled air directed onto the transformer to prevent the transformer from overheating. A short time later an electrical contractor was called in to replace the transformer with one that handled a higher electrical load.

[@ComputerHistoryArchivesProject]

Hi Bloqk-16, you make an excellent point. These giant machines were massive energy hogs especially for businesses that had no prior experience with big computer systems. Vacuum tub based equipment, electromechanical tape drives and disks, A/C, all drew a great deal of electrical power. And there is the added issue of having "clean power" since early machines could not handle minor spikes in voltage very well. Thanks for raising this observation. It is certainly worth mentioning from a historical perspective! ~ Victor, CHAP

[@GH-oi2jf]

It would only be a problem if the facilities management was poor. The power requirements of each piece of equipment was known and documented. A new computer room would have been designed to provide a specified amount of power sufficient for whatever would be installed initially, and with a margin for growth.

[@stargazer7644]

These things didn't exactly plug into electrical outlets. The power supply had to be installed by an electrician.

[@jamesmacnair9528]

Real trip down memory lane for me. But this must be very early in the life of the 360. I swear that a number of the things that they showed either never made it to product or looked very different when it did finally become a product. Has anyone actually tried to figure out what was talked about but either never made it to products or looked very different when it finally did.

[@Beun007]

Awesome!

[@ComputerHistoryArchivesProject]

Beun007, Thanks very much. `

[@ForgottenMachines]

11:34 I wonder what the name of that console terminal-typewriter/printer "teleprinter" is? I guess it's an early version of the IBM 1052?

[@ComputerHistoryArchivesProject]

Hi @ForgottenMachines, Looking closely at the image, I would think you are correct. A detailed list of IBM/360 peripherals available at the time of this film may give more clues to its model number. Thanks for spotting this. Good question. : )

[@rty1955]

my first machine I worked on was a 1401, then shortly after the 360/30 with a whopping 48k of memory, that's only 48,000 BYTES of memory! I wrote in assembly Lang as COBOL was too slow and too big to run on the machine. it ran 2 partitions or processes at a time a background process (BG) and foreground (FG) process. we always would say "man, what I could do with 4 more K!" the disk drives (IBM -3211 held 5 Meg of data.) ibm called that "external storage" and with memory called "main storage" I used to patch software using the front panel blinking lights and switches, then when it worked well, I would force a crash, obtain a core dump, then correct the original code. sometimes we would use REP cards in an object module before execution. this is a lost art or programming, but what I could accomplish in assembler, COBOL programmers could only dream of. ahhhh I miss those days

[@ComputerHistoryArchivesProject]

Ahh yes, the magic of Assembler language. Only the Guru's of Tech knew Assembler.

[@Lazarus0357]

+roy Yung Same here mate! When I moved from a 1401 to a 360 it was amazing, WAAAA, what they will think next! Regards

[@Lindsay5137]

ASSEMBLER ? Crikey ANY assembler - even basic was a great advance, the first software course I went on (1964) we had to write what was basically a "Supervisor" (Operating System) in machine language, this was to teach us how that interrupt system worked. all was fine until I discovered a program check in my program check handling routine. Talk about disappearing up your own fundamental. (You are allowed to laugh)

[@newjerseybt]

The days of blowing a register after a simple program modifications (adding instructions) or causing a dump because you forgot to subtract "1" from a register before inserting a floating $ sign. Fun days.

[@Bhakti-rider]

@@Lindsay5137 Good one. Yes, I delved into channel programming a bit while I was an operator trainee after having studied four languages in a business school.

[@nicholasmaude6906]

The Solid Logic Technology (SLT - en.wikipedia.org/wiki/Solid_Logic_Technology ) that IBM developed for the System/360 series was used for the Saturn 5's LVDC (Launch Vehicle Digital Computer) that guided the rocket during launch via the rocket's analogue FCC (Flight Control Computer) which gernerated the steering commands for it.

[@livesimplyandhumbly]

Back then, they did sooooooooo much with a computer with less power than a Raspberry Pi. Scientists and engineers, remarkable people. Today, what most of the masses can do with smart phones far more powerful than the 360 is twit and watch cat videos.

[@kaioocarvalho]

Well, that is just normal. Back in that time only professional computer scientists and trained company analists had access to computers. The leigh person had no chance to ever use one to do something stupid. Besides, even though as a human activity silly videos and tweets might seem futile, computationally speaking, handling trillions of messages hurdling around the globe through billions of devices in an intelligent and reliable manner, or playing back high resolution and high framerate sharp and colour accurate video with excellent digital sound are indeed absurdly remarkable features. Take twitter and youtube on a time machine and I bet they would prefer to use your phone.

[@johnazhderian5734]

The Havard Mark I has 1,600 bytes of RAM. The IBM 360 has RAM of 8,000 bytes to 8,000,000 bytes (8Mb). (8:10)

[@goldgeologist5320]

Used one in college. Have things improved!

[@txna58]

First computer I ever worked with.

[@CARLiCON]

question: once they had developed secondary storage on magnetic tape, why did they have to keep using punched cards? couldn't they just use tapes to store the programs & data? like early PCs, e.g., Apple II, Commodore, TRS-80 used cassette recorders?

[@ComputerHistoryArchivesProject]

Hi CARL iCON, good question. Punch cards and punch card machines were very prevalent at that time period. Raw data was input from punch card operators (data entry) by hand, to create the data on the cards. The cards were read into the computer system and processed, with output to tape, or directly from card to tape. I don't think one could type fast enough to input directly from keyboard into a tape machine. Cards were used less and less over time, as other faster ways of data and program input were developed. Perhaps someone else can come up with a better explanation. Thanks again for your question. ~ Victor, CHAP

[@CARLiCON]

@@ComputerHistoryArchivesProject Thanks Vic, I think I see what you mean now. So before tape, there was only punched cards for secondary storage of programs & data, therefore the infrastructure was all in place to have multiple keypunch operators keying in large volumes of data that way. The keypunch machines were not connected to the mainframe, (only the card reader) so there would have been no way to get that data into mainframe RAM (to be saved to tape) other than the programming console which would have been much much slower. Thanks for the excellent video!

[@GH-oi2jf]

Hollerith cards were used for input because they were easily edited by adding or changing cards.

[@NJMerlin]

The 360/20 was not, I believe, part of the original announcement, and the 360/25 absolutely wasn’t. Early 360 literature mentions the 30, 40, 50, 60, 62, and 70. (The 60 was dropped, the 62 was replaced by the 65 before shipping, and the 70 by the 75.)

[@regiirecords8829]

Looks like a scene from "2001"- fashion and all. 🤘🤖

[@rameynoodles152]

Damn! Up to 340,000 bytes (characters) per second for a TAPE DRIVE! WOW! That's still impressive! A CD drive at 1x speed is only 150,000 bytes per second.

[@jzero4813]

Sean Ramey Well, standard LTO tapes now transfer at 150 MB/sec, so maybe 340kb/sec is not exactly impressive *today*...but still very impressive for the time.

[@rty1955]

Sean Ramey tapes were much faster than disk as far as I/o operations

[@stargazer7644]

These tape drives wrote data in parallel a byte at a time across multiple tracks on the tape unlike today's serial implementations.

[@laustinspeiss]

It’s interesting to see AFAIK, the 360 was the transition from Big Anything to ‘Big Blue’…. The later System 360s shifted to the iconic blue frames, and stayed that way.

[@NJMerlin]

The 360 came in blue, red, yellow, or gray, but almost everyone got blue unless they had more than one system In the same room. The 370 added white, and the 4300 series added green.

[@thomasanderson1416]

Waaay ahead of its time, but how to run it?

[@miles2378]

How was the transistor conected to the printed contact wires on the substrate?

[@ComputerHistoryArchivesProject]

Hi Christopher, perhaps this may help: see Wikipedia's article on IBM's Solid Logic Technology en.wikipedia.org/wiki/IBM_Solid_Logic_Technology

[@FORTRAN4ever]

In my early days as an applications programmer when working in the early 1980's at an installation in a Sperry Univac 1100 series environment, mass storage (that is, disk storage) was available at a premium. A lot of large files were written out to magnetic tape. Fast forward to the early 1990's, Fast forward 20+ years, when working in an IBM mainframe environment (mid-range IBM 9370 in a VM/SP OS), still mass storage (in the IBM world called DASD or Direct Access Storage Device), the systems programmer who was the gatekeeper of mass storage was still stingy with allocating disk storage even though there was ample storage available. It was a real pain.

[@stargazer7644]

I worked on Univac 1108s.

[@Syncopator]

I miss the 2741 terminals. I learned to program on them in the early 1970s. APL -- it seemed at the time to be the programming language of the future-- what it turned out to be was explicitly designed for low-baud rate printing terminals. And without such a terminal, it's a bit of an anachronism.

[@JohnSGruber]

I loved it too. I remember the domino operator could solve simultaneous linear equations. When you saved the workspace you saved all of the variables values too --so you could resume from exactly where you left off.

[@paulstubbs7678]

What is the terminal at 14:48 - 14:54 ? An IBM terminal with a round display - that must be their first, but I have been unable to find anything about it.

[@ComputerHistoryArchivesProject]

Hi Paul, good question. I did a little research on the IBM site, didn't find an answer yet. Perhaps one of our IBM'ers can point us to more info on the round display terminal you mention. IBM had many, many products in the early days that came and went so fast there is not a lot of info on them, without some digging. Thanks for spotting this! ~ Victor, at CHAP

[@paulstubbs7678]

@@ComputerHistoryArchivesProject I've dug through a few early IBM manuals etc, no joy. Also Google searches for IBM terminal. display station, '360 peripherals etc. etc. same, no joy. I saw it quite a while ago in a different video, just one image, and thought it might have been an early prototype that never made it to market - however being in this system 360 video makes me think it was more than that.

[@paulstubbs7678]

@@ComputerHistoryArchivesProject IBM 1015 Inquiry Display terminal Finally found it - yah - and has this been a loooong search. threader.app/thread/1173829224026628101 - half way down.

[@ComputerHistoryArchivesProject]

Hi Paul, fantastic! Yes, you found it! Thanks. Now we have the model number, we can find more info too. Here's some more ed-thelen.org/comp-hist/IBM-ProdAnn/1015.pdf ~ Victor, CHAP

[@paulanderson79]

I worked on a 4381 in early 1990.

[@rty1955]

paulanderson79 then you must have run DOS/VSE XP. The 4300 series were meant for DOS. MVS ran much slower than DOS on these machines

[@int53185]

Pretty soon Intel and AMD won't be able to downsize circuitry any farther, then we wil enter the molecular or optic computer age. Making silicon obsolete.

[@paulanderson79]

I've said for many years that electrons are too big and the speed of light is too slow. Thus both impact upon computing power and speed.

[@Mr.Pop0]

Damn, my office coulda looked like damn star trek in the 60s

[@ironchef3500]

I love this stuff

[@ComputerHistoryArchivesProject]

Me to. ~ CHAP

[@ddostesting]

I wonder if a large core storage like this exists anywhere in the world today.

[@ComputerHistoryArchivesProject]

Thanks for your comment. Check out this link for the NSA's giant computer data center. en.wikipedia.org/wiki/Utah_Data_Center

[@Lindsay5137]

Yep, I've got an NCP from a 3705.

[@stargazer7644]

I had a customer who ran this technology until 2000.

[@alexpowers3697]

Did they really cut up the floor to hide all the wiring from module to module? There's not one outlet or wiring shown in the pictures--marketing or real? I wish my office were that clean...

[@lookoutforchris]

Raised floors are standard in most high performance computing centers. Not sure they do raised floors in your average data center but the big boys still roll like that.

[@alexpowers3697]

Sweet--thanks for the info.

[@jameshay7247]

The floor is elevated, the cables run underneath. That was the standard set-up.

[@z84c00]

Do not belive these stories about "raised floors" and "cables underneath the ground" - IBM used an advanced wireless power and data transfer technology recovered from the Roswell alien spaceship. Later then they revealed what we know now as "WLAN".

[@martincox4520]

Moving equipment involved a lot of planning to get the holes in the floor tiles in the right place

[@NJMerlin]

That mock-up of the 360/30 doesn’t even resemble the earliest mock-up I ever saw.

[@johnlister]

I've always loved how IBM's brochures (and it now seems sales promo films) had red accents on the equipment. And there were no-cost RPQs for a variety of colo(u)rs. Does anyone know anyone who had a colo(u)r different from IBM blue?

[@johnlister]

And here we are 55 years later and there are programs that were compiled for System/360 that will still run now. A great achievement.

[@Bhakti-rider]

@@johnlister Banks, insurance companies, government agencies...

[@五十嵐秀雄-k3d]

私は日本人ですが、IBM360システムの保守をしていました、故障すると、日本全国のむ、機械の保守をしておりました、とても、なつかしく、思います、その後、NECのACOS1000の開発、および保守を担当してきました、とても懐かしく思います。

[@ComputerHistoryArchivesProject]

Google Translate says: " I'm Japanese, but I was doing maintenance on the IBM360 system, and when it broke down, I was doing maintenance on machines all over Japan, I feel very nostalgic, and then NEC's ACOS1000 development, And I've been in charge of maintenance, I miss it very much." ---- Thank you for your lovely comment! ~ Victor, CHAP

[@leonardoantonio8756]

They have pretty good use at Banks, It was a clone of this system I remember, The Century 200 was pretty much like the IBM360

[@minorshan]

So, how many characters equals a MB?

[@ComputerHistoryArchivesProject]

Hi Shannon. Well, generally, 1 character in plain text uses 1 byte of memory. There are 1,048,576 bytes in a megabyte, so this would be about 1,048,576 characters. Approximately 500 pages of double-spaced text. Some older computers used different byte sizes... perhaps someone else can comment more based on their experiences.

[@SkuldChan42]

So when he emphasizes 8 million characters of memory - is about 7812 kilobytes of ram, or 7 megabytes ((8000000/1024)/1024) - which for 1964 really is a colossal amount of memory really.

[@cyrex686]

Depends on ASCII or ANSI, I'll just pretend it's character = 1 MiByte They were talking about peripheral transfer speeds of up to 1.2MB/s, that's faster then some cheepo USB drives...

[@m1lkb0n3z]

I believe for System/360 IBM used its own encoding system, known as EBCDIC, for "Extended Binary Coded Decimal Interchange Code". It did use 8-bit bytes, though, with 1 character (two hexadecimal "digits") per byte. A kilobyte would then be 1024 bytes (by convention, the nearest power of two to 1000 was used), and a megabyte 1024*1024 = 1,048,576 bytes. Arithmetic could be done directly in Binary Coded Decimal, and this was used for financial calculations, to eliminate rounding errors. IIRC, the Intel processors used in the 1982 IBM PC (model 5150) had a few instructions for doing BCD arithmetic. Contrast with Control Data mainframes from the 1970's (I think they were designed by Seymour Cray before he left to start his own company), which used 6-bit characters, with an encoding they called "Display Code". This is because the machines used 60-bit words (because 60 has so many divisors: 2, 3, 4, 5, 6, 10, 12, 15, 20, 30), so 10 characters could be packed into a word. Digital Equipment (DEC) systems found it convenient to describe data in octal rather than hexadecimal, due to the data formats they chose.

[@rty1955]

Milk Bone imho hexadecimal was so much better than octal. I always chuckled as DEC for using octal. I read core dumps for a living back then. Some days I would have piles of dumps in my office to explain to various programmers what went wrong. I debugged COBOL, PL/1, DL/1 and various subsystems like CICS and POWER. I could make wall paper out of my IBM certificates of bug fixes.

[@franklhota5019]

OK, I'm sold. I'll get a 360 as soon as it gets a port of Google Chrome.

[@bertoid]

I've actually personally owned several IBM System/360's, back in the 1980's. Got them for almost nix, and trashed them for the gold. I wish I'd kept the front panels though.

[@jasongomez5344]

12:10 Aren't these actually System/370 pictures?

[@ComputerHistoryArchivesProject]

Hi Jason, there were numerous models of the System/360, so those pics were probably a later model. The film was made by IBM before the System/370 was built. Pretty sure they got that photo right. But thanks for the comment! ~

[@jasongomez5344]

Oh right. I thought they might have been some of the spliced-in content that I see in some of your restored films. Cheers.

[@GH-oi2jf]

I think the photos are of early 360 prototypes. They don’t look like the 360 or the 370.

[@SteveMacSticky]

I wonder what operating system they ran

[@ComputerHistoryArchivesProject]

Hi Steve, good question. The IBM System/360 Operating System was known as "OS/360", a batch processing operating system developed by IBM for the System/360 mainframe announced in 1964. It was partly influenced by the earlier IBSYS/IBJOB and Input/Output Control System (IOCS) packages for the IBM 7090/7094 and others. Also known as Disk Operating System (DOS). There is a lot of history in the IBM mainframe O/S world. Fascinating research, if you are so inclined.

[@SteveMacSticky]

@@ComputerHistoryArchivesProject thanks, very interesting

[@Bhakti-rider]

They had a choice... there was DOS, OS/PCP, OS/MFT, or OS/MVT. Many MVT installations also used HASP to spool jobstream input and printed output.

[@ShainAndrews]

I"m sold. Where do I purchase such a mahvulus machine?

[@ricardopereiradeazevedopb1973]

I start with a IBM-PC XT at 8MHz and 640K RAM + 360K Disk Drive, but, this video is the Begins !!!

[@GH-oi2jf]

No, there were many computers before the IBM 360 Series.

[@litoboy5]

cool

[@RWZiggy]

They had me at one thousand character per second paper tape reader, I gotta get me some of that!

[@ComputerHistoryArchivesProject]

RWZiggy, me too. But how much paper tape would be needed to hold the KZbin video database? ....Charles, CHAP

[@RWZiggy]

@@ComputerHistoryArchivesProject Well asking such a question to old computer geek you'll cause an answer to be generated! Since one inch 8 channel IBM tape had ten "characters" to the inch, and youtube database is 45 TB, we have 4.5 tera-inches. Though usually one bit was used for parity which I think is good idea for so many inches of data, so we'll take 8 / 7 of that for 5.1 tera-inches of paper. A roll was 1,000 inches so a mere 5.1 billion rolls of paper. I've only used paper cards back in the 1980s, my mom used the smaller 5 bit teletype tape on a "word processor", one tape with people's data like name and address, and one tape with form letter to send them.

[@ComputerHistoryArchivesProject]

Thanks! You've got some good experience here! It is interesting to remember how many different formats of paper tape and cards there have been over the years too! ~ Victor

[@supercompooper]

Back then, Memory referred to storage :)

[@joshuamathews5507]

28 people are too sissy to write any low-level code. Real programmers use punch cards and tape drives.

[@MJK1965]

🎶We've come a long ways, baby!!!🎶

[@MasterMindmars]

Very beautiful. But nothing about the specs of the processor.

[@ComputerHistoryArchivesProject]

Hi MasterMindmars, yes, I see. I believe the processor specs did vary a lot with the model type and how it was configured (and there were many models). The big System/360 Model 75J (NASA used at least five of these in 1966), only had about 1 MB of main memory each, to start with. Its amazing how much a megabyte cost back then. ~ Victor

[@310McQueen]

Oh my, how far "solid logic technology" has come. Way more primitive than I remember when I worked in a fab in 2004, which was 15 years ago. Even then the wafers were 8" or 12" I wonder what they are now!

[@axizepp]

it depends on the process. if some power/analog components are manufactured they probably still use 12" wafers.but when it comes to cpu's they use 30"

[@tedtech5340]

Wow I had no idea that networking existed in the 60's at all

[@Lindsay5137]

Yes it did

[@stargazer7644]

We've had national data networks since 1861 when Western Union built its first transcontinental telegraph line.

[@braintrustbroker9255]

August 7th 1944 Mark One Specs: 1600 bytes of internal storage. Input output versitility. A modem. Backward compatibility. AND LATER: April 7, 1964 System 360 Specs: Massive tape drives. Punch cards. CRT Displays. Teletype interface. And a pretty girl to operate it (Yes the first programmers were women.) 8 Mega bytes of internal magnetic core storage, no refreshing, no power needed to maintain memory. Solid Logic Technology: the grandfather of the first IC Chips. (Yes the IC makers were women.) Optional Multi-processor systems. Paper Tape readers, high-speed printers, Removable disk pack hard drives. AND NOW: January 7th 2020 a wearhouse full of these computers can be replaced by a single $5 Raspberry Pi zero W in your pocket. WOW! Now that's progress children. Now I am waiting for Superman Crystal Optical Holographic AI computers to hit the marketplace.

[@randywatson8347]

This is so cool

[@Beun007]

So, this was actually the cradle of the digital computer as we know it today...

[@GH-oi2jf]

The System 360 was not the cradle. Computers were fairly mature in concept when the 360 came out.

[@Scuud52]

7:19 Fitter, happier, more productive

[@alphonsocarioti512]

I've got more computing power than that "wall" in my pocket now!

[@rolfjohansen5376]

8 mill characters of core storage , well that should en sufficient to write a letter to grandma ...

[@retrocysper3709]

13:41 Internet?

[@ComputerHistoryArchivesProject]

Hi Retro Crysper, When the narrator said "on line," he was not meaning the Internet. Back in the sixties, "on line" access usually meant the computer had access to the data on a disk or an active resource, instead of "off line" meaning data stored on magnetic tapes that someone had to find and load onto the system. Good question! Thanks. Victor at CHAP

[@allanegleston4931]

@@ComputerHistoryArchivesProject when i worked @ # old honneywell, we had a teletype line to headquarters back east . we had to get all of our jobs done and on the line to them by 1500 or 3 pm as they closed 3 hours earlier.

[@EnronnSierra]

But why didn't humanity improve? There is still warm, crime, violence, stupidity. This sounded like it would have organized society and make it a better, equal place for everyone.

[@knightwatchman]

You're yearning for a utopian society. It will never exist because humans are greedy. If people can do it, they will exploit technology, as you will see with AI, to gain power and wealth.

[@honkhonkler7732]

EZ, humans do what humans do... Use new technology to devise more efficient means of killing each other.

[@ebridgewater]

The audio is so quiet 🔉

[@oldi184]

International Business Machine!

[@majordenisbloodnok7048]

Commonly called "I'm Bl**dy Marvelous" behind their backs by us users