Slashdot Mirror
Terabyte Drive to Debut Later this Year
mytrip writes to mention the news that Hitachi will be releasing a terabyte storage drive this year. "These large drives also will get incorporated into televisions and personal video recorders. Hitachi, among others, already sells TVs with integrated hard drives in Japan and other markets. While large drives start out expensive, the price drops relatively quickly. Computer makers pay something in the 30-cent range for a gigabyte when buying hard drives, Healy said. The price at retail is around 50 cents or less."
Some of these much-slower latency discs should catch up and overtake hard-disks. http://en.wikipedia.org/wiki/Holographic_Versatile _Disc mentions "a demonstrated maximum of 3.9 TB for 3 micrometer separation on a 12 cm disc."
Finally we can start backing up our entire hard disks. Even these new ones!
I could've sworn that...oh, that's right: http://hardware.slashdot.org/article.pl?sid=06/07/ 30/2124225
Anyone else think the comments just weren't rendering right before they turned off ABP and saw ads?
...but at not quite 0.91 TiB, I couldn't help feeling gypped if I bought one of these.
We're starting to reach the point where hard drives are so large we're not sure what to put on it. Well, lots of people will; they're a boon for people doing video editing and they'll keep you in episodes of the Sopranos for months. But drop one into a regular desktop PC and your typical average user simply won't be able to fill it up; or if they do, they'll already have reached a point where they don't know what 80% of the data on the drive is.
As a sometime hardware tech, I'd really love to see the manufacturers using some of this capacity for redundancy, rather than sheer space. Run the drive as a RAID unit, with each surface being one "drive", and use two of the surfaces for parity. You'd lose up to 40% of the capacity of the drive but it would become much more reliable. Sealing off the platters from each other might take up so much space you'd lose one platter, and might mean more expense since you'd need multiple head units, but again, the reliability would improve enormously. While this still isn't quite as reliable as having multiple separate drives RAIDed together, it would be convenient and transparent to the user, and make dead drives (mostly) a thing of the past.
> why is it hard to find SCSI drives in these high capacities?
SCSI/SAS/FC drives typically spin at 10k or 15k RPM, compared to 7.2k RPM for ATA drives. The higher rotational velocity means more work to keep the heads on track, so the data densities aren't quite as high. Higher rotational velocity also causes more aerodynamic turbulence at the platter edges, which can make the platters vibrate. Most enterprise 3.5" disks actually use 2.5" platters in order to keep the disk edges farther away from the case to minimize the turbulence.
The enterprise class drives also devote more of the platter area to error correction. If you look at the uncorrectable error rate for enterprise class vs. desktop class drives, you will see about a 10x higher uncorrectable error rate for desktop class drives.
So smaller platter, lower density, and more bits for error correction are the primary factors which cause enterprise drives to be lower capacity than desktop drives.
-- Chris Caudle
Not the ones that I have seen. There are basically two main failure modes on a hard disk. Either the bearings on the motor give out, or the reserved area for mapping out bad sectors fills up and you see bad sectors. Controller failer is *much* rarer than either of these two events. If you ask me controller failures are more likely to be down to people not taking proper ESD measures.
Onboard flash caches and larger ram caches are going into the next generation of hard drives. Other than that, nothing much is going to change in the near future.
When the OS is aware of the flash and ram caches on the drive, it will instruct the drive as to what to cache so when the computer is started up next time 50% of the boot code is in the flash and starts running very quickly while it loads the rest of the boot code into ram and feeds it out. Beyond that there isn't much the hard drive can do differently to speed up normal use unless you parallel more platters (which raises heating, noise, energy use, and weight of the servo arm (which slows it down)). In most cases it's better to use several drives in a RAID configuration to obtain the same benefit. You should also consider getting a system that can support 8GB of RAM and loading it up with fast ram so it never has to page to the hard drive. Unless you use photoshop. Then you're out of luck - RAID is as fast as you're going to get.
It's going to be a bumpy start, but flash caches will significantly speed up the hard drive during boot up and a few other times.
The limiting factor is the speed of the mechanical parts, and you can only get very tiny incremental improvements in speed for each large improvement in the mechanical structure. So they are pursuing other methods to raise the apparant speed.
-Adam
Will there ever be an upper limit to hard drives? I know we just started using perpendicular technology, but there must be some kind of physical limit to the platters. Another question is why is it hard to find SCSI drives in these high capacities? Or at least in newer SAS drives.
From what I've read over the past year, perpendicular recording supposedly will offer densities somewhere between 2x and 5x over existing longitudinal recording methods. That puts 3.5" SATA/IDE drive somewhere in the range of 1TB to 2.5TB before they hit the wall again. For 2.5" SCSI, 600MB up to 1.5TB. I suspect that things will top out around 3x-4x densities over existing drives. (GMR longitudinal recording was supposed to bring us greater gains then it did. You can look back at the original announcements of bit densities and then look at what finally made it to market at the high end.)
SCSI drives are a different form factor. They use smaller platters inside to allow for higher rotational velocities (10k/15k RPM) and faster seek times. That limits their capacity per platter.
(I did all the math about 3 months ago for another article, looking at existing bit densities vs what perpendicular recording bit densities were estimated to be at the upper end.)
Wolde you bothe eate your cake, and have your cake?