Happy Birthday, hard drive!

October 12, 2006 11:20 AM

The hard drive celebrates its 50th birthday Timothy Prickett Morgan reviews the history of the hard drive, introduced to the world in September, 1956 as the IBM 305 RAMAC . Imagine life without the hard drive, without the ability to store and quickly access bootlegged MP3 and video files and pr0n large data collections. (To anticipate, yes, Mr. Morgan may know the history of technology, but firearm nomenclature, perhaps not so much.) Also Tom's Hardware Guide interviews Seagate's Senior Field Applications Engineer, Henrique Atzkern, on the hard drive's future.
posted by mojohand (19 comments total)

I'm using the hard drive of the future today! And I've got to say, it's quite nice.
posted by me & my monkey at 11:30 AM on October 12, 2006

Have you seen MicroSD cards? I got a device that uses them a few days ago, so I'd ended up with a 1GB card to go with it.

As I held it in my hand (that's actually a misnomer, it rested comfortably on my fingertip) I couldn't help but to think how strange it seemed to have the equivalent of 4x my very first hard drive (a 200MB SCSI disk in an Apple Performa 575) in the form of a tiny chip.

Despite people cussing them out when they fail or incorrectly draw the ire of the poorly informed (a ton of people that I've encountered refer to the entirety of the computer itself as 'the hard drive'), I, for one, am thankful that hard drives allow me the ability to store thousands of songs in my pocket and the million and a half things I've got stored in my 1TB array back home.
posted by owenkun at 11:37 AM on October 12, 2006

My first hard drive was a 20 MB MFM half-height 5 1/4" model. It rocked -- no more swapping floppies!

Now I have a networked 1.5 TB array full of avis mp3s pr0n personal data.
posted by LordSludge at 11:37 AM on October 12, 2006

What I want to know is why, after all these years, we're still mucking about with hard drives. Shouldn't have we gone solid-state a long time ago?
posted by Afroblanco at 11:37 AM on October 12, 2006

I'm pretty sure modern hard drives aren't Winchester-style any more, either; they all park the heads off-platter (with a voice-coil actuator instead of a stepper motor, it's easy to do this automatically).

The famous IBM Deathstar problems were due to the head landing on (or taking off from) the platters instead of staying in the special head-parking slot.
posted by hattifattener at 11:41 AM on October 12, 2006

My first computer with a hard drive was a Macintosh LCIII, circa 1993. I was in eighth grade. I still remember my friend Peter Shulman (now, I understand, pursuing his doctorate at MIT) telling me, "You might think that 80 megabytes sounds like a lot, but trust me, you're going to want the 160."

It amuses me greatly to realize that, since my current computer has 160 gigabytes of hard drive space, I could in theory fit the entire contents of that old machine onto my current one one thousand times.
posted by Faint of Butt at 11:58 AM on October 12, 2006

I remember a meeting in about '81 of the Apple computer club of Washington. When discussing a 2 MB hard drive, one of the members on a panel said, "Do you really need this? Can't you just program more efficiently?"
posted by dances_with_sneetches at 11:59 AM on October 12, 2006

goodness. did anyone else, on first reading the post, figure it was a still-functioning hard drive celebrating its 50th birthday?? i'd buy one of those... but, i guess that'd be a floppy disc hooked up to a power supply hm, only several orders of magnitude larger...
posted by n y my at 12:02 PM on October 12, 2006

He who dies with the biggest hard drive wins.
posted by weapons-grade pandemonium at 12:17 PM on October 12, 2006

Shouldn't have we gone solid-state a long time ago?

That's what I linked to upthread. But they're still very expensive, and will remain much more expensive than traditional hard drives for a while. Right now, they're really only practical for the mobile computing niche.
posted by me & my monkey at 12:28 PM on October 12, 2006

no, you Atari and Commodore whiz kids from the late 1970s and early 1980s, you did not have a real computer by my standards because you did not have random access storage

Who the hell is this guy? I can forgive (and probably not notice) the warned-of firearm ahistoricity. But the above is pretty silly for a computer history article isn't it?
posted by freebird at 12:44 PM on October 12, 2006

I wonder if the hard drive has a future. The theoretical max bandwidth of SATA is 1.2 Gbits/sec.

Verizon is currently offering Fiber to the home that is capable of a downstream data bandwidth of 622 megabits/sec (even though they are limiting the packages they sell to 50 megabits/sec).

802.11n has a bandwidth of 200 megabits/second.

Granted these speeds aren't as fast as the theoretical bandwidth of hard drives, but at those speeds it may not matter. 1080p HDTV needs about 75 megabits per second. We may come to the point that you just store everything on a central store (less prone to viruses and crashing drives) and access it over one of these high speed networks.

I suppose these data centers will need hard drives, but we as users may not.

Then you get into the thorny question of if you buy something from itunes, would you actually get a copy of the song transferred to your directories on the central data store, or would you just get a pointer to a single master copy? Wouldn't that be more efficient?
posted by Pastabagel at 1:10 PM on October 12, 2006

My first hard drive was a 30MB MFM (unusual at the time: generally 30MB was RLL, a different kind of encoding that took better hardware.) It was 5.25" wide, and double height, and it sounded like a jet engine spinning up on powerup. I still have it, although I no longer have a controller card that can talk to it.

I remember being rather proud that I could fit my entire computing 'life' into 5MB of disk space... everything I needed to use the computer routinely. Games changed all that, and then music, and then video.

In 1986, even my wildest flights of fancy wouldn't have imagined the machines I'm using in 2006.
posted by Malor at 1:34 PM on October 12, 2006

Tom's Hardware Guide: So magnetic storage will be the main storage medium! Do you think canmight last another 10, 20 years... or even longer?

Atzkern: Magnetic storage devices will absolutely be the most important storage type.

Well, yeah. Right up to the point when solid state quantum optical storage comes along offering multi-terabytes per cubic inch.

He who dies with the biggest hard drive wins.

From exhaustion? I'd rather die with the smallest, densest HD. I don't miss my twenty pound 5.25" 5meg.
posted by loquacious at 2:38 PM on October 12, 2006

I used to covet the 1mb hard disk drive for the commodore 64. It was way out of my financial reach, at the time, though.

Later, at my first real job after college, I installed a "hard card" into the IBM PC in the accounting office. It was an easy way to install a storage device to the dual floppy system. (It was better than swapping out the floppies for the customer data.) 10 MB would hold all of the customer data my boss could ever want.

Later, when I moved to a real IT gig, I worked on a IBM System36, with the hard disk drives as big as a car tire. And then to the AS/400 which still had pretty huge disk units... about the size of a medium pizza. 4 of them in the system rack, were as tall as my leg. I'm forgetting the disk capacity.

Then we made the jump to a PC network from the dumb terminals. And we spend a small fortune on a 5gb raid ... $30k, easily. 5 gigs for a firm of 100 people. We'd never, ever run out of disk space!

I have 10 times that on my PC now.

I imagine in another 20 years, I'll have 100x that on my cell phone.
posted by crunchland at 2:59 PM on October 12, 2006

Ten years ago, I paid $200 for my first ONE GIGABYTE (OMG HUEG) IDE hard drive, and that was *at cost* from the PC builder I worked for.

Now, that $200 buys me FIVE HUNDRED gigabytes. I carry around one gigabyte of storage on my keychain. My cell phone has half a gigabyte of storage on a chip the size of my fingernail.

I have more bandwidth to my house now, for $17.95/month, than the ISP I worked for ten years ago shared amongst its customers.

Technology is amazing.
posted by mrbill at 11:37 PM on October 12, 2006

"I'm pretty sure modern hard drives aren't Winchester-style any more, either; they all park the heads off-platter"

Nope, ramp-on/ramp-off heads are actually fairly uncommon in desktop drives; they have a landing zone on the platters which is coated with lubricant, unless this has changed in the past few years. I'm sure some do; IBM were trying to use it to differentiate themselves from the competition around the time of the Deathstar, but it's not really much to get excited about.
posted by Freaky at 1:59 AM on October 13, 2006

My first hard drive had 5 MB. It just blows me away to think of such things. Of course, there was nothing on that 5MB but a bunch of code (it was at work, in 1981) and a bit of data for testing. Now most of us have so much data in our primary storage, it can't be backed up in any timely fashion. LOL!
posted by Goofyy at 7:13 AM on October 13, 2006

One thing not mentioned about the RAMAC is that, while it had 50 platters, it only had 2 read/write heads. So besides having the cylinder seek time, it also had to move the R/W assembly up and down to select the correct platter.

The details in Morgan's article seem kind of sloppy. The 1311 held 2MB on 100 tracks at 1025 bpi, not 3MB on 50 tracks at 1100 bpi. The 3340 was 2964 rpm, not 4000 rpm. Also, the 3340 was not the Winchester. The 3340 was the disk drive. The Winchester was the removable head/disk assembly (the 3348) that would go into 3340. It had 2 heads per surface, not 2 heads total. I used to service these machines and I never heard of or saw one ‘walk across the floor’.

I am also skeptical of the 30-30 story about the Winchester name. I have seen three different stories on the web: two 30MB disks side by side; 30MB fixed platters and 30MB removable platters; 30MB disk with 30ms seek time. This last one is on IBM’s site but is stated like IBM read it somewhere else. But the 3348 was 35MB (later also 70MB) with a 25ms seek time and never had removable platters.

The biggest problem with the 30-30 idea is that ‘Winchester’ was the CODE name for the device. The reason you give a machine a code name before it is publicly announced is that you are trying to keep details about it secret from your competitors. You don’t give it a code name that says ‘I’m not going to tell you any details about my 30MB, 30 ms secret disk drive.’ I remember IBM code names like Hickory, Minnow, Gulliver, Piccolo…just random names. My personal theory on Winchester is that it came from the Winchester Mystery House, a tourist attraction in San Jose.

The 3340 did use a voice coil, but parked the heads in the landing zone, not off-platter like earlier drives.
One thing not mentioned about the RAMAC is that, while it had 50 platters, it only had 2 read/write heads. So besides having the cylinder seek time, it also had to move the R/W assembly up and down to select the correct platter.

The details in Morgan's article seem kind of sloppy. The 1311 held 2MB on 100 tracks at 1025 bpi, not 3MB on 50 tracks at 1100 bpi. The 3340 was 2964 rpm, not 4000 rpm. Also, the 3340 was not the Winchester. The 3340 was the disk drive. The Winchester was the removable head/disk assembly (the 3348) that would go into 3340. It had 2 heads per surface, not 2 heads total. I used to service these machines and I never heard of or saw one ‘walk across the floor’.

I am also skeptical of the 30-30 story about the Winchester name. I have seen three different stories on the web: two 30MB disks side by side; 30MB fixed platters and 30MB removable platters; 30MB disk with 30ms seek time. This last one is on IBM’s site but is stated like IBM read it somewhere else. But the 3348 was 35MB (later also 70MB) with a 25ms seek time and never had removable platters.

The biggest problem with the 30-30 idea is that ‘Winchester’ was the CODE name for the device. The reason you give a machine a code name before it is publicly announced is that you are trying to keep details about it secret from your competitors. You don’t give it a code name that says ‘I’m not going to tell you any details about my 30MB, 30 ms secret disk drive.’ I remember IBM code names like Hickory, Minnow, Gulliver, Piccolo…just random names. My personal theory on Winchester is that it came from the Winchester Mystery House, a tourist attraction in San Jose.

The 3340 did use a voice coil, but parked the heads in the landing zone, not off-platter like earlier drives.
posted by MtDewd at 7:24 AM on October 13, 2006