Slashdot Mirror
Hitachi Predicts 3D Hard Disks by Year's End
daria42 writes "Hitachi has announced that its perpendicular, or 3D, hard disks should be out by the end of 2005." From the article: "Today, hard drives record and store data in a longitudinal fashion, with the read/write heads scanning over a horizontal plane. In perpendicular recording, data bits are aligned vertically, allowing for more data to be squeezed into a finite area. Put another way, data will go from being stored on a two-dimensional XY grid to living in a three-dimensional XYZ space."
I'm tempted to say: "Nothing new move along"; but I appreciate that it's quite different when applied to digital media.
Although storing information in 3D is nothing new, that's how you get music in stereo from a vinyl record.
The data transfer rate would be around eight times faster...
Current affordable hard drives are, well, pretty farking big already. I care a lot less about capacity now than reliability. I despise hard drives and look forward to the day when they are just a bad memory. (And I haven't even been burned badly, since I back up.)
Give me a guaranteed 5-year lifespan on a drive, then you'll have my patronage... more gigs don't get my attention anymore.
I'm just amused this was modded as insightful.
:P
True enough I s'pose
"This is your life, and it's ending one minute at a time."
That's a nice theory and all, but that isn't what this drive does. There are plenty of links scattered throughout this article with decent explanations.
Your data is safe during a defrag, you need make no sacrifices.
Defrag works by taking a fragmented block of data and making a copy somewhere else on the disk. Once the copy is made and verified the inode is updated. If the power goes during the copy it doesn't matter because the inode still points at the original data.
I've never fully understood what impact on-drive caches have on this kind of situation though, so maybe there is a small chance of losing small amounts of data.
You always hear about forensic people being able to get at data that's been written over many times on a hard drive. I wonder if there's any way that could be built into a hard drive. Could you store multiple sets of data in the same place on the hard drive and reread it at different sensitivities, or however they recover overwritten data?
Let me know, hard drive experts.
Meh. If you want "high" capacity, you have to invest in the infrastructure to support it. SATA isn't about speed (it's not - the drives are about 0% faster than IDE), but about being able to support increased densities of drives in an enclosure due to simplifed cabling. Buying a 3ware or Highpoint card and a 5-in-3 enclosure just gets you to a place where capacities most desktop users would call staggering are utterly normal.
The other aspect of your wish for higher capacity drives is that your expectation of capacity will increase over time as well. If you have drawers full of IDE drives now, you're going to want them in the future, whether the drives inside are 1TB or 5TB. That's just the mindset of folks who buy multiples of large hard disks.
Five years ago people were ripping their whole CD collection to some compressed format and storing them on a PC. Now people are storing the uncompressed audio, and some people are moving their whole DVD collection onto PCs. In another 5 years people will be moving their collection of HD discs to their PCs. The need for multiple drive storage systems is not going to disappear just because you hit some arbitrary capacity.
While I'm at it, I might as well point out that data distributed across several spindles is inherently safer than data on a single spindle. This is why I stick with midrange drives rather than the largest models I could buy (250GB instead of 400+GB). Backing up a huge single drive is a pain in the ass if you don't have a BIGGER single drive someplace else (er... or a tape, I guess. A lot of tape).
-- I wanna decide who lives and who dies - Crow T. Robot, MST3K
"Works to avoid" does not equal "avoids"
I ran a news server on ReiserFS for about three years and a crappy XML db that creates seven files for each piece of database in the db for about two years. ReiserFS worked ok for about a year or so, then it started getting very slow. I think it was due to fragmentation since moving the files to a different drive them moving them back fixed the problem. Of course, our use was an extreme example (500Gbyte drive that was full of mostly 1k or much smaller files), but it was still an issue.
how do I visualise this? Data in jelly blubber with a read/write needle swimming through it? Data gets read out where two laserbeams cross?
Actually, what you describe exists. There's a team that was making, a decade ago, transparent gelatinous cubes containing bacteriorhodopsin, a light-sensitive protein similar to the light sensor in your own optical rods in your retina.
By indexing the cube with two different lasers simultaneously, you could cause the bacteriorhodopsin in an indexable 3D location to switch between two different conformations (foldings), or read fluorescence which indicated its' current configuration. Thus storing a rewritable bit in a small region of a 3D transparent cube of "jelly blubber". Data did in fact get read out "where two laserbeams cross".
I think the rapid growth of HD sizes, coupled with the fact that you have to keep the cube moist, is why they've not managed to make a marketable product yet. (Incidentally, allowing the cube to dehydrate would make the data unreadable. But, it didn't destroy it. If you rehydrated the cube, you could get the data back... which is kinda cool...)
Here's a link to
an early description of the technique from 1996.
I stole this sig from someone cleverer than me.