Slashdot Mirror
IBM and Fuji Announce Tape Storage Breakthrough
robkill writes "IBM and Fuji have announced a breakthrough in the amount of data that can be stored on magnetic tape, a 15X improvement to 6.67 billion bits of data per square inch. IBM estimates that it will be 5 years before this hits the mass market"
is a lot of porn.
What?
Registered Linux user #421033
When a 500 GB hard drive costs $75, can be thrown across the room and have a chance of working, weighs the same as a tape and can be easily inserted/removed in bulk with software management and barcode readers to keep track of it all for you.
Until then, tape will stick around. I have a feeling it might be a while.
Here's a better article.. They're claiming 8TB before compression on an LTO-sized tape. Tape record smashed with 8 terabyte format
You've clearly never worked in an enterprise-level environment. Larger companies like WalMart measure their storage requirements in petabytes. To these companies, the physical space required to store tape backups is significant, especially when you consider some records (such as employee records) have to be kept for up to 30 years after the employee leaves the company. Plus, virtually all accounting records must be kept for a large number of years by law.
FYI: Sony claimed 11 billion bits per square inch quite some time ago.
It's always good news when someone figures out how to store more bits into the same amount of space, and I'm sure that companies like IBM and Sony will keep pushing the limits.
If it's for tape, it depends on how long and wide it is. They give the denisty, so you can compute how much a particular tape size would hold... If it is about 6 Gbits per square inch, and is made into half inch tape that is ten feet long, then it'll be about 60 square inches, which is about 45 GBytes. If it is about 750 feet, which seems pretty realistic, then it'd be something like 3 TBytes per tape.
I've probably scrwed up the math, but I'm sure you get the idea.
Which is just as good, the sheer size obtainable using tape drives is just mind boggling.
On a side note, this article wasn't just light on details, it was shockingly devoid of all technical details as to how this was acheived. At least this article mentions the new density is acheived with a new tape medium coating.
Sheesh, the linked story might be interesting to stock-market droids, not slashdot readers.
You can't legislate goodness. Let each to his own destiny, by will of his freely made choices.
So if you combine the tape density breakthrough with the Linux device driver breakthrough, can you go faster than the speed of light? Or does SM just need a thesaurus?
-h-
There's also some advantage in separating the storage medium from the read/write heads. If either part in a hard drive fails, you're literally fscked (except for some really expensive recovery solutions by Ibas or the like). On the other hand, you can always put an optical disc in a brand new drive. And if a disc is scratched beyond readability in your current drive, chances are you can read it with another drive in the future.
Escher was the first MC and Giger invented the HR department.
All you naysayers, understand that we had exactly the same sort of announcements before the Linear Tape Open (LTO) standard was developed. IBM led a group of manufacturers to develop a standard built around a few breakthroughs in tape density and drive head technologies. They predicted 10X (or more) capacity, 5X (or more) throughput, etc. and it would be available in 5 years or so. Sure enough, LTO-1 came about and immediately led to a tape storage boom. Quantum pushed DLT to about its limits, Storagetek upped the ante with their very high speed formats, etc. Everything got cheaper. Tape stayed relevant. I predict the very same trend in the near future...
There exists no way of exchanging information without making judgments. --Bene Gesserit Axiom
Upgraded YOUR tape drive lately? DDS has to be the slowest "modern" tape formats in existence. I'll never touch one of those drives again. Worthless. Try VXA-2 or M2.
-matthew
"THERE IS NO JUSTICE, THERE IS ONLY ME." -Death
In my experience, the trouble is not in writing lots of data to tape, it's in reading it successfully afterwards. /only half-joking
Recursive: Adj. See Recursive.
In the 1950's, data was recorded on 7-track tapes at 200 bits per inch. Those 7 tracks recorded 6-bit characters (EBCDIC?) with one bit of parity. A 2400 foot tape might hold some 5 million characters (upper case only, please). By the mid 1960's tape densities more than doubled to 556 bpi.
Sorting algorithms were written to sort information using mag tapes; the speediest would make the tape vibrate in the vacuum columns with a minimum of reel-to-reel motion.
By the 1970's, most shops changed to 800bpi, 9-track tapes, which would happily handle 8-track encoding. Then came 1600 bits per inch -- you could store an amazing 50 megabytes onto a single tape.
There was a constant temptation to compress data so as to stuff as much onto the tape as possible. As a result, many graduate students earned their assistantships by decoding tapes written with oddball parity, density, and encoding combinations.
The scattering matrices from my dissertation are encoded onto 9-track 1600 bpi tapes, carefully stored in my climate uncontrolled attic.
I need it for all the pictures of your mom.
Yo momma so fat, I need all that storage for just one picture of her.