The History of Computer Data Storage, in Pictures

Nowadays we are used to having hundreds of gigabytes of storage capacity in our computers. This was pure science fiction only a few decades ago. For example, the first hard disk drive to have gigabyte capacity was as big as a refrigerator, and that was in 1980. Not so long ago!

Pingdom stores a lot of monitoring data every single day, and considering how much we take today’s storage capacity for granted, it’s interesting to look back and put things into perspective. Here is a look back at some interesting storage devices from the early computer era.

The Selectron Tube

The Selectron tube had a capacity of 256 to 4096 bits (32 to 512 bytes). The 4096-bit Selectron was 10 inches long and 3 inches wide. Originally developed in 1946, the memory storage device proved expensive and suffered from production problems, so it never became a success.

Selectron tube
Above: The 1024-bit Selectron.

Punch Cards

Early computers often used punch cards for input both of programs and data. Punch cards were in common use until the mid-1970s. It should be noted that the use of punch cards predates computers. They were used as early as 1725 in the textile industry (for controlling mechanized textile looms).

Punch card Fortran program
Above: Card from a Fortran program: Z(1) = Y + W(1)

Punch card reader and punch card writer
Above left: Punch card reader. Above right: Punch card writer.

Punched Tape

Same as with punch cards, punched tape was originally pioneered by the textile industry for use with mechanized looms. For computers, punch tape could be used for data input but also as a medium to output data. Each row on the tape represented one character.

Punch tape
Above: 8-level punch tape (8 holes per row).

Magnetic Drum Memory

Invented all the way back in 1932 (in Austria), it was widely used in the 1950s and 60s as the main working memory of computers. In the mid-1950s, magnetic drum memory had a capacity of around 10 kB.

Magnetic drum memory
Above left: The magnetic Drum Memory of the UNIVAC computer. Above right: A 16-inch-long drum from the IBM 650 computer. It had 40 tracks, 10 kB of storage space, and spun at 12,500 revolutions per minute.

The Hard Disk Drive

The first hard disk drive was the IBM Model 350 Disk File that came with the IBM 305 RAMAC computer in 1956. It had 50 24-inch discs with a total storage capacity of 5 million characters (just under 5 MB).

IBM Model 350, the first hard disk drive
Above: IBM Model 350, the first-ever hard disk drive.

The first hard drive to have more than 1 GB in capacity was the IBM 3380 in 1980 (it could store 2.52 GB). It was the size of a refrigerator, weighed 550 pounds (250 kg), and the price when it was introduced ranged from $81,000 to $142,400.

Really big hard disk drives
Above left: A 250 MB hard disk drive from 1979. Above right: The IBM 3380 from 1980, the first gigabyte-capacity hard disk drive.

A hard disk drive called SyQuest was targeted for personal computers, and for many years had no relevant competition in terms of transferring large desktop publisher documents. The first SyQuest SQ306RD introduced in 1983 had endless (at the time) 5MB hard drive for most of the available data types, also audio and video. In 1986, a 44 MB SQ555 and SQ400 were available on the market.

Above: Removable hard disk SyQuest 44 MB.

The Laserdisc

We mention it here mainly because it was the precursor to the CD-ROM and other optical storage solutions. It was mainly used for movies. The first commercially available laserdisc system was available on the market late in 1978 (then called Laser Videodisc and the more funkily branded DiscoVision) and were 11.81 inches (30 cm) in diameter. The discs could have up to 60 minutes of audio/video on each side. The first laserdiscs had entirely analog content. The basic technology behind laserdiscs was invented all the way back in 1958.

Above left: A Laserdisc next to a regular DVD. Above right: Another Laserdisc.

The Floppy Disc

The diskette, or floppy disk (named so because they were flexible), was invented by IBM and in common use from the mid-1970s to the late 1990s. The first floppy disks were 8 inches, and later in came 5.25 and 3.5-inch formats. The first floppy disk, introduced in 1971, had a capacity of 79.7 kB, and was read-only. A read-write version came a year later.

Old floppy disks
Above left: An 8-inch floppy and floppy drive next to a regular 3.5-inch floppy disk. Above right: The convenience of easily removable storage media.

Magnetic Tape

Magnetic tape was first used for data storage in 1951. The tape device was called UNISERVO and was the main I/O device on the UNIVAC I computer. The effective transfer rate for the UNISERVO was about 7,200 characters per second. The tapes were metal and 1200 feet long (365 meters) and therefore very heavy.

Old tape drives for OLD computers
Above left: The row of tape drives for the UNIVAC I computer. Above right: The IBM 3410 Magnetic Tape Subsystem, introduced in 1971.

And of course, we can’t mention magnetic tape without also mentioning the standard compact cassette, which was a popular way of data storage for personal computers in the late 70s and 80s. Typical data rates for compact cassettes were 2,000 bit/s. You could store about 660 kB per side on a 90-minute tape.

Compact cassette and Commodore Datasette.
Above left: The standard compact cassette. Above right: The Commodore Datassette is sure to bring up fond memories for people who grew up in the 80s.


DECtape was a magnetic data storage introduced in 1963 used mainly with Digital Equipment Corporation computers. It was designed to be durable and reliable enough to act as a computer’s operating system main storage medium. The tape was 0.75 in (19 mm) wide and was accommodating 6 data tracks which were paired with non-adjacent tracks. It could have been written/read in any direction, and it was almost error-proof. The DECtape was used on Digital Equipment’s PDP-series of mainframes and minicomputers, and each tape could store 184K 12-bit PDP-8 words or 144K 18-bit words.

Above left: Dual DECtape unit for a DEC PDP-11. Above right: DECtape removable magnetic media.

There are so many interesting pictures from “the good old days” when you look around on the web. These were some of the best we could find, and we hope you liked them.

If you are interested in the history of computer science, you should check our Gallery of Early Computers and old tech reviews from BYTE Magazine.

Picture sources:

The Selectron. The punch card. The punch card reader and writer. Punched tape 1 and 2. UNIVAC magnetic drum. IBM 650 computer magnetic drum. The IBM Model 350 Disk File. 250 MB hard drisk drive from 1979. The IBM 3380. Laserdisc vs DVD. Held Laserdisc. 8-inch floppy drive. 8-inch floppy in use. UNISERVO and UNIVAC I. The IBM 3410. The compact cassette. The Datassette, SyQuest, TU56DECtape, DECtape I and II.

And as always, Wikipedia was a great source for checking out the actual facts.

EDIT: We removed the comment about the Commodore Datassette sound, since it was a factual error. We also removed this: “(For those who weren’t there, you could hear the sound of the data being read as a high-pitched, screechy sound while you were loading your programs.)” We also went with some of the comment suggestions and decided to add paragraphs regarding the SyQuest and DECtape.


Can’t get enough of nostalgia? Don’t miss: The history of PC hardware, in pictures.


  1. Actually you could not “hear the sound of the data being read as a high-pitched, screechy sound while you were loading your programs”.

    That only happened if you tried playing the cassette in a regular tape player.

  2. Interesting that you chose to include the Selectron vacuum tube, because it’s the only technology listed here that was not for secondary (permanent) storage. Selectrons were used for primary data storage, like a modern computer’s RAM. Other interesting forms of data storage include bubble memory, magnetic core memory, twistor memory, and thin film memory.

  3. Kreiger, you’re absolutely right. Fixed. The writer was struck by a whisp of senility, and old memories of a Spectrum with a regular cassette player.

    Jim, you have a point there. Still it’s such a cool device that we included it anyway.

  4. The interesting thing is how quickly and efficiently progress in computiatonal and storage capacity has evolved through a tacit public (governmental) and private industry partnership. Think of what might happen if an equivalent effort was made to revolutionize energy production, storage and economical usage.

  5. Thanks for the memories. The floppy-booted Amiga especially hit home. I remember borrowing a friend’s 20MB hard drive for my Amiga 500 to run a BBS. It was brand new — this huge, heavy, triangular thing (to match the “sleek” contours of the Amiga) that plugged into the side.

  6. Actually, you could hear the tape loading on my old Spectrum, although, if I remember correctly, the sound became muted shortly after the programmer loader was loaded and the data had commenced.

  7. For the record, you could hear the lead-in carrier track on Atari cassette tapes. Once it started to load, you’d get a friendly “bluuuurp” sound with each block read. If the tape was bad and the computer lost track of reading, it would start screeching horribly again. It was always a great diagnostic for bad tapes, because you could hear it go south when you were in the next room, waiting 15 minutes for a program to load.

  8. Thanks for the history. It is amazing that there are many as/400s still running banks across the country.

    Old tech is still pretty new. If aliens unearthed all of our tech in the future the older stuff would be easiest translated. Perhaps someone should carve a rosetta stone for computer technology!

  9. I worked for Equifax in the 80s. They unplugged their last punch card around 1985. One day a beginner system programmer accidentially destroyed the VTOC (Volume Table of Contents) for the disk drive that housed the operating system boot. So the systems were down for quite some time. The experts at IBM were called and they were struggling to get the system back up. Then a retired Equifax guy was called in. He had the system up in 30 minutes. He retrieved a punch card unit from storage (saved to donate to a museum), and started the system from a stack of punch cards. LOL

  10. Maybe you couldn’t hear the screeching of cassette data loading on a Commodore machine, but you could on the Atari 8-bit machines.

    I think the compact cassette system must have been the slowest and least reliable computer-mass-storage scheme ever tried on any significant scale. I had floppies that worked fine decades after they were made, in spite of heavy use. Those cassette tapes sometimes wouldn’t even last a week.

  11. I was just at the Computer History Museum in Mountain View a few days ago and saw a couple more obscure storage technologies I’d never heard of before:

    One was delay line memory, which Keith mentioned above. It stored data as pulses in a tube of mercury. A 1-meter tube could store about 1000 bits. The UNIVAC I had 10 “memory tanks,” each holding 18 tubes. (

    The second was the IBM 1360 Photostore, which used an electron beam to write data onto plastic chips. It was the first device that could store a terabit of data. (


  12. Using the BBC Micro or Electron microcomputer system you would get the tape noise from cassette-based program loading at least until the first piece of the program loaded. The computer had relay control of the tape deck to turn the volume off an on. This was with the “official” BBC tape deck.

    Chris Morgan

  13. Marc – Not only did the Timex Sinclair 1000 (ZX81) use a cassette tape, so did Coleco’s ADAM. Of course they screwed up with the whole not using a standard cassette tape (the through holes where in different places)..

    Now to see if I got out bid on my TS-1000 on eBay.

  14. David: If I had to guess that looks like a breadboard. Toss up a pic of the other side. Would certainly help both myself and anyone else trying to figure out what you have there.

  15. You skipped over my favorite piece of archaic ram. The Williams tube. It was used in UC Berkley’s Wirlwind machine. It was essentially a cathode ray tube where each pixel was a bit of memory.

    Programing using Williams tubes was tricky because reading a bit destroyed it. Also if the memory wasn’t continuously refreshed it degraded in seconds. Wirlwind was later converted to slower but much more reliable magnetic core memory. Slower ram actually provided a performance boost because there was no longer any need to refresh the memory every few seconds.

  16. In the early 1960s, my uncle created one of the first business data service firms, that processed primarily paychecks, etc. (Think of ADP.) He and his partner had no computers, but rented the second shift on the MetLife computer (where they worked days.)

    Of course, they picked the most modern, forward-looking name they could think of “Punched Card Services.”

    Later, while a grad student at Yale, we had an early IBM 360 (Model 44.) It had several multidisc storage units that looked like small washing machines. When they were not in use, the techs would make use of their cooling capabilities by storing their lunches in them.

  17. You neglected to mention that since 1980 vast amounts of data were stored on computers (esp. monitors)with the use of Post-it notes.

  18. Excellent post!

    About the datassette: the model pictured in that photo didn’t have speakers, but other models did; on those, you could hear that screechy high-pitched sound (dialup-modem like) of the data being transferred. In fact, the speaker was indispensable to fine-tune the playback head when data in some old or low-quality cassette refused to load.

    BTW, for those visitors interested in articles about computer history written in spanish (hard to come by), I’m posting a series here:

  19. I was fortunate to have worked with everyone of these; except for the selectron tubes.
    I started out working for ISS/Univac making the magnetis drums and went on to work at General Electric where I became an expert at all other others.
    I went from GE to Space Systems/Loral (also known as Philco Ford and Ford Aerospace).
    I was fortunated to have participated in the transformation from Mainframe to Servers, while I was the Superviser of OPS at Space Systems Loral.
    My youngest son, who is getting ready to graduate in an IS field just sent this to me and I was thrilled. He learned alot from is article.
    Thank You for sharing.

  20. The pegboard was called a control panel in my IBM world. They were mostly used in “pig iron” machines, tabulating machines (402,403,407), card punches (514), interpreters (557), etc.

    They were used to manipulate to manipulate the card data… move fields, direct data to registers, selectors, counters, …etc.

    The 2848 Control unit for 2260 displays used delay lines. It was a wire delay line, with input and output transducers. The delay lines were kept in an oven at a constant temperature. The length of the wire was critical, since there was no synchronization of the bit stream. You expected the input bit to come out at a pretermined time. If my memory serves me right, a 960 character display needed 4 delay lines to store the data. The data in the delays lines was used to prime a character generator that was a magnetic core device and the cores were actually visible within the plane as the characters the they were going to generate… I always wanted one of those.

    Good memories….

  21. Not to pick or anything, but the Commodore 64 was the beloved computer on which I cut my coding baby teeth all those years ago, so I just couldn’t let this one slide. You could, actually, hear the sound of the data being loaded from the datasette, however, it was only audible if you were using a TV as your monitor, and turned the sound all the way up — I’m guessing it was some sort of interference with the RF output sent to the TV. I used to use this to help debug assembler programs too, as the RF noise from the CPU was also audible in this fashion, and you could actually *hear* what your code was doing. (High pitched whine? Your code’s in a tight loop.)

    Hope this helps. 🙂

  22. Look closer at that paper tape — It has seven bits plus a small sprocket hole. The Univac 1218 I programmed was an 18-bit computer. Each row on the tape had six data bits and one parity bit. It took three rows for one 18-bit word.

  23. I’m pretty sure you could hear the tape sound on the TRS-80 Color Computer, too.

    Also, you claim that tapes did about 2,000 b/s; there may have been some that went that high, but the ones I remember were either 300 (Commodore) or 1500 (Color Computer).

  24. There was one interesting memory system you missed it was a 3850 mas storage. it had spools of magnetic tape that downloaded info that was downloaded to them and a kind of a bee hive that the spooles of tape were put in. So when some one needed the info on the tape a servo device would go to the location in the bee hive ge the tape which was on a spool an put it in a tape reader and down load the info to a disk drive for use.If the customer had alot of data the storage for the tape casets could be added making one very long machine.places like Sears,Allstate,Bell labs and one of the other insurance had them I found it a wonderful system to work on.

  25. The TRS-80 tape cassette drive actually had a setting which allowed you to listen to the sound of the data as it read it.. Oh the memories!

  26. Hi Folks;

    My former computer repair instructor put a link to your site on his student website, so here I am. Fascinating stuff. I remember the old Commodore computers and owned two of them early in my “geek” days, and have had several that used the larger floppy drives as well. One of my best friends learned programming in the 1970s with punch cards and Fortran. I’m sending him a copy of this article as he’ll probably enjoy the portion on the Univac. I believe that was the computer he was trained on. One of my earliest computers was a Timex Sinclair. Remember that one? And I played around on a CPM machine a few times … yep, that article brought back LOTS of memories …

    Lynn Magnuson
    New Orleans, LA.

  27. What about the Kado (not sure of the spelling) I rember my Dad running one in the Early 80s’ it had a 8″ floppy disk. He used to brag that he could run his entire business on 2 floppy disks!

  28. I love the fact that the HAL 9000, with its ability to recognize the movement of lips and human drawings, was running off of punch cards. For those of you who don’t believe me, see “2001” and fast forward to the point where HAL detects the fault in the AE-35 unit. Dave will ask for a hard copy of the report, and HAL prints out an actual punch card.

  29. Here’s another person who remembers that the old Atari 410 cassette recorder for Atari 400 and 800 computers did play the amazing screech erroneously attributed to the Commodore. I’m fairly confident that the only reason the screech existed was because you’d surely think the system had hung if you didn’t hear something. At an effective read/write rate of about 35 bytes per second (600 baud with long inter-record gaps), reading or writing a file took forever. for substantiation.

    1. The Atari cassette recorder was stereo, one track was intended for audio and fed to the TV speaker, the other was for data. Most commercial games were recorded in mono, or used the audio track for anti tape-to-tape copy protection exploiting head alignment errors, and I seem to recall that the Atari itself recorded data over both tracks.
      The chip that decoded the tones to data had no reset pin, and the OS used it in whatever state it powered up in, so data transfer wasn’t as reliable as it should have been. It was possible to reset the chip by manipulating some of its registers, I wrote a program that did that and could load data at much higher speed, although I could never get the Atari to reliably save above 1407 baud, other than by manipulating tape speed.

  30. Hey Keith! I came to the comments to make the exact same comment about mercury delay lines being my favorite…and I’m named Keith too. 🙂

  31. Back in the mid to late 1960s, I worked on the Litton DIANE system on the Grumman A-6A Intruder. Our support bench was called a SASE, and was driven by a paper tape/aluminized mylar tape system. When we had a correction to the program, we got a slip of tape from Litton, and cut and “patched” the correction into the tape. Today we are still “patching” our systems. Also in the DIANE system, there was a drum memory. It sat between the B/N’s (Bombardier/Navigator)legs. If I recall, it was 144 tracks, and we had to do two fetches of 4 bits, to make a one 8 bit, byte. We didn’t even have ICs then, a gate was built “chord wood” style and was one inch square on each end, and about one and a half inches long.

  32. I recall a description of a system that used a video camera pointed at an oscilloscope to store 32 bits altogether.

  33. My dad just turned 70 last week and when my roommate showed him her iPhone he was amazed (he’s not real techie) It got him talking about hen he was in the Air Force back in the 50s and early 60s they had whole BUILDINGS for just one computer that couldn’t do what a pocket calculator can do.

    I worked in a hotel several years ago that had a mainframe computer that took the 8 inch disks that had a whopping 76K on them and used tapes to run the back-ups. Seeing the pix of those “floppies” brought back a lot of memories.

    Great article

    1. I worked on an IBM air defense computer system starting in ’61.  56,000 vacuum tubes, 2500 miles of wire, 1 million semiconductor diodes.  We used punched cards, magnetic tape, magnetic drum memory, and the system had 65,536 words of 32-bit core memory.  I have been in the computer industry since then and have lots of great memories of the changes in technology over the years.

  34. David: Looks like a control panel from some Unit Record device. Similar to those for the 402/408, but probably a one-off machine. Wired a lot of machine in early 60’s.
    BTW: If you tuned a radio between stations and set it on top of an IBM tape drive, you could create music on it by the data you sent to the tape. Knew a chape who spent months to make it play Mary Had A Little Lamb.

  35. Many devices used SHIFT REGISTERS as a storage media.

    For example, the early display terminals (“CRTs”) were in fact teletype terminals, operating at blazing speeds of 1200..9600 Bauds, connected to COM1 of mini-omputers and early PCs [*]. The display area typically had 24(/25) lines, 80 chars/line, for a total of 1960..2000+ characters (25th line: status data) which needed to be stored somehow.

    A shift register is like a bench, a row of chairs. Put many of these around a clock until the end of the last bench touches the begin of the first one. You now have a circle, a ring with thousands of chairs arranged around that clock.

    The way it works, on every chair sits either a “0” or a “1”. Each time the clock ticks they all stand up, make one step to the right and sit down again. The last one on every bench now sits in first place of the next bench and…

    The electron beam, as it is redrawing the display advances to the same clock beat. After every 7th tick it gets to see who sits on “Bench No.1” and displays the graphic that corresponds to the 7 bits sitting on this bench at that instant. Say new data arrives, bit by bit it is entering the circle at some “Bench Nr.77”. As new bits enter bench-77 the ones leaving bench-76 are lost as they step on a trap door and fall into the bit bucket.

    [*] The alternative was memory mapped video, a portion of the PC’s RAM used as video buffer and a graphics card (CGA = IBM/PC Color Graphics Adapter), converting the RAM content into RGB video signals, sent a video monitor (a fast TV without tuner/receiver circuitry).

  36. I still have some “floppy” floppy disks (5.25 inch) & a 40 Meg disk that occupies twice the space of a modern hard disk. State of the art technology, that was!

  37. May I contribute with a missing landmark in the history of magnetic tape data storage? Its the DC100 Data Cartridge, a 3.5″ digital data and program mass storage unit developed to fit into thee HP9825 desk computer, launched in 1976 with a 16-bit “triple core” hybrid microprocessor – fantastic for numerical processing, interfacing and automation and years ahead of the 8-bit microcomputers from APPLE and IBM PC with 5.25″ floppy disk drives.

    The quarter-inch-wide tape minicartridge (QIC) & drive is much smaller than the 8″ floppy disk & drive from the 70ties (3.5″ floppies appeared in 1981) and much faster and reliable than analog storage on cassete tape. The QIC originated an entire tape-backup industry and evolved to capacities of up to 20-Gbyte. QIC-backup dominated the scene till surpassed in capacity and speed by the hard disks in the 21st century.

    More about:

    Application examples of the HP 9825 (with the DC 100) from our lab: and

  38. Hi. The relay on the Electron and BBC Microcomputers was to control the cassette motor, so that it stopped after loading or saving. It may also have controlled the volume, but only with specially-adapted cassette recorders. Hearing the data was essential!
    I remember getting my first 5.25″ floppy drive, single-sided, 40K. It cost about £100.
    I also remember using the cassette relay to switch my amateur radio transceiver between transmit and receive. We used to send programs to each other by radio. Magic days!

  39. That’s a nice piece, back in 87′ when I took computer classes to learn BASIC (I was in seventh grade), we were required to get a 3.5″ floppy disc. I got one too though never used it and I still have it. It’s an antique, Wonder what I’ll get for it on eBay. 🙂

  40. I remember seeing a product that would print programs on a laser printer that could be read by a hand-held plug-in device. The output looked similar to the optical data storage area on the back of some states drivers licenses. I don’t remember what it was called but I actually did spring for one. Programs could be entered in a blazing three to five minutes instead of hours!

    This made getting programs for those Apple ][ computers much easier as we used to type in whole programs in BASIC right from sources such as Nibble Magazine. ( It was not uncommon (among nerds at least) to spend several days keying in a multi-page program that looked cool. The rest of the world just wondered what we were doing.

  41. I remember the cassette tape drive of my first computer, Radio Shack’s TRS-80 Model 1 Level 2. It’s hard to believe how many cassettes I used up learning that system. Later on I got a TRS-80 Model III with a 5.25″ floppy disk. Of course as soon as I could I created flippies for it, to save money and double my space. The orginal was only 180KB, by making another eye hole in the floppy I could have another side to the disk. Later years I went to a school that still had some of the Model IIIs in use. One had an external hard drive, about the size of three laptops stacked on top of each other. It held a whole 5MB of data. It ammazed me at the time.

  42. interesting trip down memory lane. I have 3 old lugable computers sitting in my garage collecting dust. combined they have about the storage space of my iphone

  43. Regarding the tape sound – You had the correct phenomenon, but the wrong computer. The secondhand Atari 800 I had as a kid used to be noisy as hell when you would load or save a tape (CLOAD or CSAVE), though it sounded more like a robotic truck horn with a chest cold.

  44. Re: EDIT: Removed the comment about the Commodore Datassette sound, since it was a factual error: Removed this: “(For those who weren’t there, you could hear the sound of the data being read as a high-pitched, screechy sound while you were loading your programs.)”

    This is actually technically correct, at least for a short period of time…

    Before the Commodore branded tape drive came out, they sold an adapter for use with an ordinary cassette tape recorder. It was a wire that plugged into the Vic20 or C64 motherboard and then into the mic input of the recorder. You then pushed play and record to start the tape drive and then ran the “save” command. It was a REAL pain because you had to get the record volume just right. While it was recording you could hear the squeals and hisses. The dedicated tape drive was really an improvement– especially because of the counter. With that you could put more than one program on a cassette tape.

    Yes.. I had both, the cable and the Commodore branded tape drive. That was a long, long time ago, in a galaxy far, far away…

  45. I started selling the 8″ and 10″ SMD Drives. Remember CDC,Fujitsu. Emulex corp was one of the first to create the 3rd party disk market selling up against Dec. We made emulation controllers that put third party disk and tape products on the manufacturers cpu. This was back in the early eighties. EMC and Netapp were not even a dream yet. EMC made 4 meg add on memory cards. They could not even spell storage. The days of the big iron wars. We were selling the stuff for about $300.00 per meg.

  46. Way to make me feel old, guys. I actually used [i.e. programmed with] both punch cards and paper tape. It felt high tech at the time. We would write batch programs using IBC job control language, and then wait for the programs to run. It seemed high tech at the time, now it seems so old. I am very glad you are keeping a historical record of computing.

  47. Great information, but one error, and one omission:

    The first floppy disk was the IBM 2315, which was 12″ across, in a hard plastic case, and held 1 Mb. It was the main disk drive on the IBM 360, Model 44, which I programmed in college.

    You omitted the DECtape, which was about 3″ diameter, has a 2″ spindle (center hole), and was 1″ wide. The DECtape’s claim to fame was it was almost error-proof. You could unroll a bunch of DECtape; wad, crinkle, & fold it up in your hand; and it could then be read and written without errors. The DECtape was used on Digital Equipment’s PDP-series of mainframes and minicomputers (PDP-8,10,11 that I know of).

    But great post!

  48. actually on some systems it would play out loud when loading a cassette, most systems used the output of the cassette player which disabled the internal speaker, i had a ti/99 that would play the cassette’s while loading them.

  49. A couple of interesting thoughts partially relating to memory and older computer systems.

    DO NOT RUN with a box of punch cards. If I was in a hurry to get my time on the computer, I always ended up tripping and spilling the lot. They then needed to be reordered and my slot was lost.

    Sound from a C64 only when listening on an ordinary cassette player OR when correctly set, typed and spaced a friend had silent night on a dot matrix printer.

    I still have Microsoft Pascal Compiler manual and disks in original plastic box.

    My first introduction to Basic programing was on the states first mainframe, accidentally created an infinite loop. Put it out of commission for about an hour, while we tried to remove it. Never saw bureaucrats move so fast!

  50. Other forms of storage which could be added:

    The 12″ floppy disk – 1Mb, hard-shell case, used on late-60s, early 70 IBM machines, among them System 360, Model 44

    DECtape – used in a number of DEC (Digital Equipment Corp. machines in the PDP series. This was a 1″ reel, about 3-4″ in diamter, held 350 Kb (if I remember correctly), and could be read/written in either direction. Its main claim to fame was that you could unspool some tape, wad it up in your hand, straighten it out (leaving folds & wrinkles), respool it, and read it without errors. Sort of like a tape-era USB key.

    The IBM 2311 (7 platter) & 2314 (14 platter) removable disk pack with completely unshielded platters, about 14″ in diameter. Used in IBM System/360 and /370 mainframes as primary storage.

  51. I recall that drum memory was also used for primary data storage (RAM). It was used as random access memory in early computers. I think the Jargon file has an anecdote about the first programmer who optimized his code to the spin rate of the drum so that the next instruction would be read just-in-time, without delay (a whole revolution of the drum, which was a considerable amount of time then).

    On the Commodore 64 Datasette many confuse the sound produced through the sound chip when loading some programs using a tape turbo (data compression program to reduce the time to load the main program), which some commercial game compilations used. Other home computer casette units may have produced sound, but on the Commodore 64 it was purely an annoying cosmetic touch by the programmer of the turbo loader (together with a flashing and/or striped screen).

    The reason for the flashing screen was purely practical though, as most commercial tapes were recorded at such a low level (to make it difficult to just dub the tape to thwart piracy) that you had to adjust the datasette read head with a screw driver to be able to load the damn game; if the screen stopped flashing while loading you knew you had to adjust the head and try again.

  52. Commodore released a new version of the casette unit, called Load-It! if I recall correctly, with a knob to adjust the head (and a led meter showing the level of signal), instead of just hole for a screw driver because of the unreliability of loading non-pirated software. I never had a Load-It Datasette, so loading games was a pain in the backside.

  53. actually on some systems it would play out loud when loading a cassette, most systems used the output of the cassette player which disabled the internal speaker, i had a ti/99 that would play the cassette’s while loading them.

  54. Hey, no one has mentioned the other storage medium that was widely used by consumers beginning around 1900 — Piano rolls. These were similar to punched tape, but were large paper rolls that each contained a tune. The mechanism in the piano used air pressure through the holes in the paper that caused the wires inside to be struck by the felt hammers to produce the tune.

  55. Interesting. First, I could never figure out why anyone would want to spend so much money to build such a big harddisk with only 5MB of space!
    Though thanks to IBM and all the genius that keep working on this, and we get to enjoy our small size laptop with Gigabytes of spaces. =)

  56. I remember my first program, on Hollerith cards. My first computer had 64 MB of memory. I thought I was living in the fast lane. Today, life really is fast and getting faster. My children know more about computers than most graduates did when I finished. Don’t ask when, I won’t tell.

  57. Can anyone tell me what the acronym TD stands for in TD machines, having to do with the old paper-tape transmission of computer information back in the early 1970s. Thanks.

  58. If you are willing to include program storage as well as data storage, how about the plugboards we used to program unit record machines? They were also used as supplementary storage in computers like the 305 RAMAC shown above.

  59. Don’t forget the magnetic card systems of the 1960’s. RCA had their RACE (random access card equipment) and NCR had their CRAM (card random access memory). Then there was the infamous (within IBM to this day) Datacell. In my collection of magnetic media I once used I have a couple of RCA RACE cards and two complete IBM Datacells.

Leave a Reply

Comments are moderated and not published in real time. All comments that are not related to the post will be removed.required