SanDisk WORM SD Card Can Store Data For 100 Years

CWmike writes "SanDisk has announced a 1GB Secure Digital card that can store data for 100 years, but can be written on only once. The WORM (write once, read many) card is 'tamper-proof' and data cannot be altered or deleted, SanDisk said in a statement. The card is designed for long-time preservation of crucial data like legal documents, medical files and forensic evidence, SanDisk said. SanDisk determined the media's 100-year data-retention lifespan based on internal tests conducted at normal room temperatures. The company said it is shipping the media in volume to the Japanese police force to archive images as an alternative to film. The company is working with a number of consumer electronics companies, including camera vendors, to support the media."

Comments

That's what they said about CD-Rs

[Anonymous Coward]

.. then they started to rot at 3-5 years, in my experience..

Post this again in 100 years, until then, it's just more bullshit marketing.

Re:That's what they said about CD-Rs

[Jason Earl]

Exactly, I would be curious to know what sort of "room temperature" tests can tell how reliable something is going to be in 100 years.

Re:That's what they said about CD-Rs

[poetmatt]

I'm sure they mean accelerated aging tests, but I have no idea if they really are applicable for real world scenarios or just good for research. Maybe someone with a bit better scientific background can comment on such.

Re:That's what they said about CD-Rs

[mirix]

I don't know if ASCII would be the best example. It's been around 50+ years, and is still readable. Hell, it's the default / only supported format for a lot of things, still. (well backward compatible extensions at least, CP-437 et al). UTF-8 is backwards compatible with ascii for that matter, too.

I'm rather disappointed with the lack of unicode support for a lot of things, in 2010. (slashdot for example).

I'd Imagine SDRSUFHC (Secure Digital Really Super Ultra Fucking High Capacity) card readers will be backwards compatible to plain old SD too. Besides, SD cards fall back to a slower plain old SPI bus, and that isn't going anywhere any time soon.

Re:That's what they said about CD-Rs

[vivian]

My CD's are working fine too - wish I could say the same about the numerous CD and DVD dives or players I have had over the years.

I really hate CDs and DVDs - the medium itself is way too easy to damage, but worse, the bloody CD drives / players just have too many points of failure in them.
I had yet another DVD drive fail on me this week - I don't watch that many movies or burn a lot of stuff, but I have gone through at least 4 DVD drives, and quite a few CD drives over the years, not to mention 3 stand alone DVD players ie. that you plug into your TV.

I still have the first CD I bought, which still plays, but the CD player in the first stereo I bought to play it in died years ago - even though the radio and tape player in the same unit work fine. I had several walkman CD players too, that have call rapped out over the years too. The CD and DVD players are just too damn flakey and prone to going out of alignment or having their lasers burn out or something. I even had one DVD burner somehow leave a burn mark on a game CD when it failed! (it created a partly melted spot on the original game CD (which has to be in the drive when playing the game) which has rendered it unusable

Anything which depends on mechanical parts that have to line up precisely for successful reading and writing is just asking for trouble, and never going to be a good long term storage solution.

The good thing about solid state storage is there are no moving parts to go wrong - so as long as the device is designed to be adequately protected from static discharge, it's going to be a lot better, in my books.

I personally cant wait to see the death of CD/DVD (or for that matter, anything involving a spinning disk) to go the way of the dinousar once and for all.

hopefully this will bring in solid state storage to replace CD's and DVD's for everything - the sooner the better.

Re:That's what they said about CD-Rs

[Anonymous Coward]

There are Archival CDs that exist, and some of mine have lasted a good 15 years so far with no errors on 100+ CDs. Only problem is, they're REALLY expensive.

For an example: http://www.delkin.com/products/archivalgold/cdr.html
$199 for 100 of these things.

You buy 10 cent CDs, you get 10 cent CDs =)

Re:That's what they said about CD-Rs

[delinear]

The knowledge to read this data isn't going to suddenly vanish. We have the technology to read wax cylinders from 120 years ago (albeit the data is often badly degraded, but these disks claim to deal with that issue) and the only reason cheap home solutions for reading wax cylinders aren't ubiquitous today are that there are very few in existence and not enough people care. If enough big government or corporate bodies have their ultra long term storage on these devices then you can be sure there will be companies offering reading services or a device to convert the data to whatever quantum state format we're all on by then.

Re:That's what they said about CD-Rs

[delinear]

The only realistic way that the things could become unreadable would be if SanDisk fucked it up and decided that some sort of uber-proprietary DRM/obfuscation nonsense was absolutely vital...

Even then, this is not data we're burying and hoping to dig up and read in 100 years (indeed, it's only guaranteed for 100 years so unless we're copying it before then we're risking losing the data anyway) - this is likely data we're going to need to access throughout that period, therefore the technology to read them won't disappear while they're still the best format. To address GP's point - the reason we can't find 5.25 floppy readers is because we don't need to - nobody is crazy enough to still be storing data this way, it will have been format shifted if it was important enough (and you're right, for "found" data that wasn't in continual use, it's trivial to build a reader - costly perhaps if the spec is ancient and the parts not being manufactured - but still trivial). The fact is, if this format is around in 100 years in substantial numbers, the tech to read it will be around, if something better comes along then critial data will have been moved to that format already and nobody except historians will care.

Oh, and spot on with the DRM issue - this is a much bigger threat to being able to read the data in 100 years than the storage media.

Re:That's what they said about CD-Rs

[delinear]

CDs have been commercially available for 28 years, strangely I don't have any problem finding something to read the data. If the format is still in use, the devices to read it will still be available. If the format goes out of fashion then you'll likely migrate your data to the new format and negate the issue (and the only people who will care in the far future - academics and historians - will have departmental budgets to build the tech to read a cache of found cards from hundreds of years ago, it's not difficult, just too costly for the average guy to bother).

Re:That's what they said about CD-Rs

[jimicus]

Hang on a minute, you sound like you know what you're talking about. WTF are you doing on /. ?

Re:That's what they said about CD-Rs

[X0563511]

I'll quote someone else. Room temperature means something very particular:

When I learned basic chemistry "room temperature" specifically meant 20 degrees Celcius. It is a fixed value.

From http://en.wikipedia.org/wiki/Room_temperature:
"For scientific calculations, room temperature is usually taken to be 20 or 25 degrees Celsius, (293 or 298 kelvin (K), 68 or 77 degrees Fahrenheit)."

Re:That's what they said about CD-Rs

[SharpFang]

Oh, of course, money-back guarantee...

So some company in 1925 sells new movie reels. The new film is guaranteed to last 100 years, money-back guarantee! You buy ten, for cost of a brand new Ford Model T.

And so, 2010 comes and you want to play back the movies. They should be good for another 15 years. But they all turned to sludge. Oh, the company is still in business, unbelievable! You even kept the receipt! So you go visit them and ask for refund. Yes, sir! Here's your $24 per reel of film, and we're sorry they failed! ...unless they are willing to insure the data for inflation-adjusted value you claim, money-back is a pathetic excuse of warranty in this situation.

Re:That's what they said about CD-Rs

[Voyager529]

I defy you to find a cheap, easy way to read 50 year old media, even if the media itself is in pristine condition. Hell, I'll even make it easier for you and set the limit at 30 year old media.

Challenge accepted: The vinyl record.

Records made 50 years ago are still readable using my Numark TTX turntables I bought last year, using the Shure M44-7 needle I bought at Christmas. I'd dare say that most records made 80 or 90 years ago - though encoded in mono - are still able to be played back presuming the media itself is intact.

Granted back then there wasn't much in the way of digital information being written onto vinyl, but there is now - it's called timecode (i.e. Serato, Traktor, Torq, etc.). So it's not THAT much of a stretch to essentially record data modulated into sound similar to an old dial-up modem on a record, then playing it back circa 2110 assuming that it doesn't spend a sunny summer day in my car.

What about barcodes printed onto paper, or some digital variant of braille? It's not necessarily the most IDEAL way of storing data for easy retrieval - in both cases the storage density is very low and thus an admittedly low capacity - but it satisfies your requirements of being a storage medium that has survived for 50 years and is still readable by hardware in active production.

100 years sounds good...

[Anonymous Coward]

Until you realize that the last reader for it will be extinct in 20.

I'll buy one so I can put it in my time capsule along with my 8" floppy and punch cards.

Re:100 years sounds good...

[phillipsjk256]

Portions of the specification are secret.

Re:100 years sounds good...

[Joce640k]

In my case that move was "yesterday". Am I doing it wrong?

I've lived through ISA/VESA/PCI, SCSI/IDE/SATA, serial/parallel/USB... nothing fits or connects any more. I really don't believe there'll be SD card readers in shops in 100 years time.

Re:100 years sounds good...

[rvw]

"Until you realize that the last reader for it will be extinct in 20."

Not necessarily. They still make turntables for LP records.
Also, if the specification is well documented, then someone can always build a reader if it really matters. File formats are likely to be more troublesome.

The LP was a medium that lasted almost a century, in a period when nothing really happened with new media. (Yeah tape, cassettes - but those came decades later and lasted for decades as well, and that's about it.)

If it really matters.... If it really matters for a big company or a government - yes. But if it matters for the average Joe Nobody, who will pay for it?

Not Enough Testing

[Anonymous Coward]

Extrapolation is a dangerous and deceptive marketing strategy. If it is supposed to last 100 years, they should test it that long.

Re:Not Enough Testing

[MichaelSmith]

Extrapolation is a dangerous and deceptive marketing strategy. If it is supposed to last 100 years, they should test it that long.

Nobody is going to sue in 100 years anyway...

Re:Not Enough Testing

[balbus000]

You don't have to wait 100 years if it fails early.

Most likely scenarios

[alvinrod]

Good for 100 years or your first fire, flood, or other natural disaster that destroys the physical media.

Also, even if these do last for 100 years, it's a certainty that there won't be any hardware left that's capable of reading SD cards. Even if there's some piece of hardware in a museum, it won't be able to interface with existing technology. Given the rapid pace of the tech industry, anything beyond 25 years is just fodder for marketing.

Re:Most likely scenarios

[Ziekheid]

I find it funny that people actually think we won't be able to recreate old technology and we would have to go to museums to get the latest working readers.
Furthermore data will just be copied and copied and copied to the latest hype so these usb cards probably won't still be around by then.

Re:Most likely scenarios

[Vellmont]

Every time I hear one of these "but.. but but nobody will have the technology to READ these things in 100 years!" all I hear is "everyone will be stupid in the future".

Someone recently created a device to read some crazy obscure technology produced by Edison to record sound on film, and that wasn't even all that valuable.

The real deal is, if the data is important enough someone will maintain the technology to read it, or re-create it.

100 years in what conditions?

[Midnight Thunder]

So they state 100 years, based on tests at room temperature. Can we assume that the media will always be stored at room temperature in 100 year period? My experience generally shows this is wishful thinking, because air conditioning breaks down, heating fails, the room is not always dark, can have direct sunlight etc. Provide me something that can last a 100 years in conditions of, at least, 30 degree centigrade variation, and then it might be interesting. Certainly I won't be around to appreciate the end results, but for archival this is a requirement, IMHO.

Re:This is going to seriously piss off RIAA and MP

[Trogre]

A large portion of RIAA's and MPAA's distributors rely on people buying copy after copy of the same media as it gets damaged or lost.

Or the shellac breaks in transit to the record store.

Re:tamper proof

[mentil]

the police can simply publish a cryptographic hash of every card they archive after they have written fabricated evidence to it.

FTFY

Re:tamper proof

[fuzzyfuzzyfungus]

Given that most courts actually believe cops whose lips are moving, I strongly suspect that the overwhelming majority of "tampering" inflicted on these cards will be done the old fashioned way. That is, there will be basically no attacks against the card itself; but the pictures taken just might be of "tidied" scenes, and the occasional inconvenient card might get tragically lost.

Sure, for some super high-profile case, the NSA can probably just 'ask' Sandisk to produce as many writable duplicates of the allegedly unique cards as they need, and have Verisign or whoever provide a 'secure' timestamp for whatever time they require. For the overwhelming majority of cases, though, that'd be overkill. Heck, the tampering would probably be more likely to cause scandal than would the existing techniques for getting the results you want. Compared to the surprisingly useless; but emotionally compelling, junk like eyewitness testimony, photographs would be practically objective, particularly if a "common photoshop artefacts detectomatic" software package can be put together so that all but the most useless defense attorneys can trivially check for mediocre hackjobs.

The practical solution

[oljanx]

We've seen a lot of discussion about "file and forget" digital storage methods. I haven't seen one that I'd trust over even a 10 year time period. The only practical solution is to periodically move your data over to the latest, long term storage medium. Make multiple copies each time, and store them in separate physical locations. I make sure to store all of my personal/financial/etc data along with family pictures and videos. I challenge you to go more than five years without wanting to watch your kids walk for the first time. This helps remind me when it's time to update.

Re:tamper proof

[Merls the Sneaky]

Get an identical card. Copy the data to HDD, tamper away, rewrite to new card.

Tamper proof my arse.