Those "IBM cards" are properly "Hollerith cards", after the inventor of the punched card system. He came up with it for the 1890 census. An operator would lower a thing with lots of pins which looked like a meat tenderizer onto the card, which had been punched with holes detailing the characteristics of a person (age, race, etc). Where a pin went through a hole and hit a contact, an electric current would advance a clock dial by one. As a result, the 1890 census was tabulated in only a year, compared to eight years for the 1880 census.

The size of the card was no accident - it is the size of the dollar bill at the time. Hollerith picked it because then the cards could be sorted into racks designed for banks to sort cash.

Yes, those IBM cards where coded in binary, but the translation was printed on the top. One side benefit of punch cards was the chad (what was punched out to make the holes) was as a random number generator of sorts. Put a bit of moisture on your finger tip, stick it in the chad bucket and out would come a nice selection of random numbers stuck to your finger. Useful for the lottery, office pools, etc. With early IBM computers (like the 1401, et al) where there was no macro to control input/output you could issue the "start read/feed" to the card reader and see how may instructions you could run before you had to issue a "read" - i.e. before the card actually hit the read head.

The IBM 701 was known as the Defense Calculator and it was announced to the public on April 29, 1952. It was also considered a Scientific Computer. The IBM 1401 was announced to the public on October 5, 1959. The IBM 1620 Model I was also introduced in 1959 and it was the first computer I learned about and used in 1962.

In playing cards, it would have been called a "52 pick-up"; in Hollerith cards, it might have been a 5200 pick-up, or worse. They used to make a metal rod for carrying the cards, so they would stay neatly in order. It looked like a short spear, or a very large knitting needle. I am not at liberty to say how I know this, but I've heard through the grapevine that occasionally college students would use the rods for medieval sword fights. Again, that is just a rumor I heard; nothing I can personally confirm.

Oh my -- the infamous 1401...

I still had to use one of these in the late '70s when attending Nassau CC at Mitchell Field (in the basement of a former barracks, no less) -- when computer sciences was still known as "Electronic Data Processing" -- oy.

They jammed constantly and the only reprieve was getting past the first programming class where we were then "privileged" to have a pool of operators key in our programs for us (anyone ever punch out an entire program in Assembler on one of these?!?!) -- woe betide the hapless victim who mis-coded their punch sheets and got their deck back with all the "O"s as zeroes & vice versa.

Rumor at the time (when some utility bills were still on these Hollerith cards) was that if you soaked them in a mixture of diluted bleach & alum they would shrink the holes just enough to pass thru a sorter unread.

I'm another USMC vet who started off with the old keypunch machine in 1967. Someone once pointed out to me that the unique thing about punch cards was that they are the only medium that can be read by both a machine and the human eye.

[Were, anyway, in the days before OCR. - tterrace]

Since nobody else mentioned it, those are Type 31 Alphabetical Duplicating keypunch machines. They were at the forefront of keypunch technology at the time, having a very typewriter-like keyboard with a separate number pad, a real numeric '1' key, and automatic card eject and feed.

JS: You were making pretty expensive confetti! Circa 1975, used punch cards sold for around $110-$125 per ton. They were very high quality paper and the recyclers liked them a lot, and they also liked green bar paper. It took about 180 boxes of punched cards (2000 cards/box) to make a ton. We financed several Physics Department parties from recycled cards and green bar paper (which I think was around $75-$90/ton, but I'm not real sure of that one).

The 701 was the predecessor to the 1401 (the first one I worked on). It was the last IBM tubed mainframe. They programmed one to control the traffic lights on Queens Boulevard in NYC. This was a first and they kept in service until the mid 90s! I did my thesis on a Univac 1600, a 20 K machine. The key punch machines never punched true. As StatPak took 18K and there needed to have room for the input data, the operators would swap out the operating system. We knew we had a bad card when the printer would start printing out paper with zeros by the box full. By the way, they still call the program that starts a mainframe the start up deck, even though punch cards have not been used in decades.

Yeah, my first IT job was about '69 as well. We were still hard-wiring unit record machines for reports and paycheck printing!

rhardin has the sorter described perfectly. Of course, there were tricks to the job, when you had thousands of cards to sort. Among them was NOT placing the follower weight on top of a stack as you added them to the input hopper. That way, the machine ran continuously, until the output bins were filling up. Problem came when you got distracted and let the input hopper run down to (almost) empty. The bottom card would often buckle, tossing the last dozen or so above it into the air, and usually damaging the card. That's when JohnBraungart's title came in!

DaveB
66 and still geeking

Here is a 1401 that has been restored to working condition; I was able to have helped out a tiny bit (no pun intended - well, okay maybe it was) with this project a few years ago. I have stood on that raised floor and listened to the glorious noise the machine makes when running a procedure that called for a lot of the machine resources at once - all the blowers and vacuum pumps and fans and motors; the smell of warm electronics and computer tape; the chatter of the printer; my my my!

JohnBraungart: I started an IT career way back when as well, and remember the 1401 as well. I suspect you mean 4K of storage (we never called it RAM); the 1401 maxed out at 16K. It was a good machine in its way, and you certainly did learn a lot operating it. After that I was "promoted" to our 7080, and then to our two 360/40 systems. After that we went modern with the 370 series, and I fell in love with VM; ended my IT stint with the same company and retired after 35 years in IT.

As a young'un growing up in the 1970s, I spent MANY hours with stacks of punch cards my father brought home. We cut them up to make confetti. The rows of numbers made it easy to cut straight lines lengthwise, then cut a second set at a 90 degree angle. The holes made it fun, akin to driving over cobblestone as the scissors went from card to hole to card multiple times in a single cut.

I was an operator/programmer during my tour in the Marine Corps. I used to HATE it when one of the pieces of equipment mangled some of my program cards, but the worst offender was the antique (even then) IBM 1401. Three units - CPU, Printer and card reader - each the size of a Volkswagen and giving a blistering 4 megbytes of memory. It was our fiscal computer and usually a lot of Marines leaving the service liked to "tinker" with the system, usually by inserting a card which instructed the machine to disregard any and all programs after a certain date (usually a month after said Marine had left the service).
This photo brought back a LOT of memories.

After assisting in the field-measuring of every floor of the Candler building in the late 80's or early 90's, its octagonal columns and other memorable attributes are hard to forget! The building started out life as a Coca-Cola bottling facility. The individual column facets are not as uniform in shape as one might imagine.

On the tab room floor around 1968

There is no backspace key on a keypunch.

Yep, that's me. I started in "computers" back in 1969, and they STILL had keypunch machines (albeit a newer, updated model from the one in this photo)at that time. Ladies known as "keypunch girls", later changed to "data entry clerks" were still the norm in '69. Not only was data created, but guys like me had to learn how to use the machine to create the cards that were read in to the computer to run the programs to process the data. Big grey tray cabinets (similar to the old card indexes in public libraries - anyone remember those?) held hundreds of trays with data. PC's were still a long way off. I worked for the Lincoln Savings Bank in Brooklyn until 1972, when I got married, and left for a better job with more money. I am still in the IT field, just turned 65, and work for Barnes & Noble, the booksellers. Haven't seen a keypunch machine in a long time, Everything now is server driven.

Great Photo! Brings back a lot of good memories, especially for us "wonks". OH BTW - they call us "geeks" nowadays.

You do card sorts by running the cards into one of ten bins based on the last number. Then stack up all the cards from successive bins and run them through again into bins based on the second-last number. After running them through that way on all the digits, they're in order on all the digits. The chief hazard is card destruction by the machinery. In 1960 the keypunchers were all women but the boss was a woman too. Gradually the self-service keypunch area grew and the keypunch service shrank, as programmers learned to type faster than the turn-around on the provided service. Today programmers are all speed-typists.

On the old IBM 029 card punch, my unjamming technique was to KICK the right side of the machine, HARD. Seldom required a second kick.

As a damp-eared U.S. Army second lieutenant in the early 1960s, I was assigned to supervise some reservists during their two-week summer deployment to the Erie Ordnance Depot near Camp Perry, Ohio. We were a supply unit there to train on the Army's sizeable bank of IBM keypunch machines and all went reasonably well for a few days, with thousands of cards
churned out to record the whereabouts and quantities of tons of military equipment. It was clear the Cold War would tilt in our direction because there was no way, I was sure, the Russkies had such technology on their side.

About five days in, a soldier whose machine wasn't punching keys correctly called me over to help him. I knew zero about these things (hardly the first time the army put someone in charge of something he/she barely could recognize) but it was clear to me that if I pushed hard on these jammed keys here and maybe that bunch there, they'd pop in place and our nation would remain safe. The army had to fly an IBM wonk (he'd be an IT guy today, of course) in from New York to fix everything.