Slashdot Mirror
Seagate Hits 1 Terabit Per Square Inch
MrSeb was one of several readers to submit news that drive manufacturer Seagate has announced (and demoed) the first hard drive to squeeze a terabit into each square inch of platter.
"'Initially this will result in 6TB 3.5-inch desktop drives and 2TB 2.5-inch laptop drives, but eventually Seagate is promising up to 60TB and 20TB respectively. To achieve such a huge leap in density, Seagate had to use a technology called heat-assisted magnetic recording (HAMR). Basically, the main issue that governs hard drive density is the size of each magnetic 'bit.' These can only be made so small until the magnetism of nearby bits affects them. With HAMR, 'high density' magnetic compounds that can withstand further miniaturization are used. The only problem is that these materials, such as iron platinum alloy, are more stubborn when it comes to writing data — but if you heat it first, that problem goes away. With HAMR, Seagate has strapped a laser to the hard drive head; when it wants to write data, the laser turns on. Reading data is still done conventionally, without the laser. In theory, HAMR should allow for areal densities up to 10 terabits per square inch (magnetic sites that are just 1nm long!), and thus desktop hard drives in the 60TB range."
Can current motherboards handle that?
Geek Hillbilly
"Seagate has strapped a laser to the hard drive head"
Well, there goes my hopes for an intelligent discussion.
"A week in the lab saves an hour in the library"
STOP! It's HAMR time!
They are on ultimately diverging paths which may coexist symbiotically forever unless one beats the other out in either cost, reliability, or functionality.
MPAA says this will cost the entertainment industry billions of dollars every year.
"Let's face it, we're not changing the world. We're building a product that helps people buy more crap - and watch porn."
Once size is sufficient, you can solve reliability through redundency.
I want a hard drive with some frikin' lasers!!!!!
If there is no God then free will is an illusion.
Not with RAID you can't. If you don't decrease the unrecoverable error rate as you increase the size of the volume, eventually you get to the point where you're almost certain to hit an unrecoverable error while rebuilding your volume. So the real question is, how is the read error rate on these tightly packed data domains?
Give me Classic Slashdot or give me death!
I wonder what the power consumption increase is if you have to strap a heating laser to the write head. Lately the market seem to reward Technology that trends toward less power usage, not more
Common Sense isn't as Common as people think...
Granted most of *us* can find something to fill it but when Dell and other bulk PC makers start including 1TB or 10TB drives in their basic PC's, most of it will still be unused by the general public. With higher MP cameras I can fill mine up with video and pics and a few converted movies/music. But with streaming options and so much available online or stored online for you, I just don't see the need to keep a ton of torrented movies and other files around taking up space and having to manage.
The more space we have, it seems the more we keep. I can see a new show as a spinoff of "Hoarders" showing just what all is in your computers HDD.
I've noticed that the more storage you have, the more junk you fill it with. At my work, we have SANs with several Terabytes of storage, mostly filled with junk. When you have millions of useless files, it becomes a tedious task to search, and backup data. In the early days, there was a lot more cleanup of stored data, and only important files were kept on disks.
-- By all means let's be open-minded, but not so open-minded that our brains drop out.
Wonder how much does the magnetic bit need to be heated before it can be written upon. When used in a laptop, battery consumption and heat generated by it might be an issue.
Can you describe what a centimeter is, relative to something else?
A fraction of the distance from the equator to the north poll?
The distance light travels in a fraction of a second? What's a second?
The length of someone's thumb, or a fraction of someones foot is possibly more relevant to humans.
that said, yes, m and cm are ok too.
but do you ask for a 8.89 or 6.35 cm harddrive?
I for one am happy to hear this. A lot of people probably thinks this will slow down SSDs move to mainstream computers but I do not. Once SSDs get to a reasonable price (under $1/GB before rebates) I can see them start becoming a common option for computers people buy in places like Bestbuy. Just imagine the "upgrade your hard drive to ultra fast speeds. Load windows faster than ever before for only $50 more" ads.
Don't get me wrong I love my SSD drives, but that tech needs to to start finding some way to move forward at a faster and a better $/GB..I guess its $/TB now :)
Power Corrupts,Absolute Power Corrupts Absolutely, leaving one person(group)in charge is absolutely corrupt.
Look forward to 60TB cellphones!
Sorry, but gray text on gray background is making my eyes bleed.
Let's bring back hogsheads and firkins!
Sorry, but gray text on gray background is making my eyes bleed.
Be SILENT Welshman or you're getting ye stones a lashing...
I am wondering what the downsides to such large density are: how fast can the laser be turned on and off? The longer it takes for it to fire the bigger the random write latency. Secondly, how long does such a laser last, can we expect 10 years from it? Thirdly, what does this mean for power consumption? More? Less? Fourth, on machines with write-heavy tasks would the drive heat up even more than they now do?
Sure, 60TB storage sounds a lot, but I have trouble believing this thing is as wonderful as Seagate makes it sound like.
I have a lot of heat coming from my rig. This oughta work fine on it!
The eternal struggle of good vs. evil begins within one's self.
I really hope they figure out how to make a faster read write speed. It is already almost a given that if you have a RAID 5 setup with 2TB drives there is bound to be an error during rebuild. There was a report on the chance of failure of an array using large drives failing during rebuild. So instead of RAID6 or are we going to have to go to a 3 parity RAID? They claimed RAID 5 should have failed in 09 and that is sata drives but if you put 60TB on a single drive you are asking for trouble IMO. Granted a nice RAID10 would be nice though. http://www.zdnet.com/blog/storage/why-raid-5-stops-working-in-2009/162
I recently did some calculations off Kryder's Law, and happened to keep the results. We would normally expect (based purely off regular continuous improvement) 6TB hard drives as early as next year, and 60TB hard drives around 2018.
So, while this is undoubtedly an improvement, it's not exactly a revolutionary one. WD et al. are probably at similar stages, either with this technology or with some other technique.
Cool. Instead of using say a half dozen smaller drives and losing some data to drive failure, now you can put it all in one place and lose it all at once!
Wuddooeyeno? IITYWYBMAD? Like nuts? eclecticallyincorrect.com
HAMR Head Sharks can hold two frickin lasers! Take that you great white hater!
I only look human.
My mother is a halfling and my dad is an ogre, so that makes me an Ogreling
As the density increases, the size of a short-stroked partition will be physically smaller too, making the seek times shorter. :)
You're absolutely right!!! Why didn't they measure it in meters??? Then I'd have some scale being able to compare it to the length of the path travelled by light in a vacuum in 1/299792458 of a second! I mean, everyone has some idea what that is...
Don't listen to those pointy-headed physicists and their ivory-tower propaganda! The One True Metre is a piece of Platinum/Iridium bar-stock painstakingly stored by the Bureau International des Poids et Mesures and roughly the same length as 1/10,000,000th of an incorrect estimate of 1/4 of a terrestrial meridian!
As an engineer I am truly impressed by the clever material breakthrough. However, I can't imagine how they can manage the mechanical precision on the positionning on the read/write locations of the HDD. Having a 1 nm wide bit size is not helpful if you are not in the same range for the relative head/disk position? Or I am missing something?
Look up racetrack memory on Wikipedia. If everything goes as planned, it'll actually out-perform SSDs (and even some DRAM) while having density comparable to hard drives.
While I suspect it'll never scale to mass production at consumer prices, maybe I'll be surprised.
The good news is that if the laser fails, the data should still be available to read and copy onto a new hard drive. If the laser was needed for reading as well, I'd be wary of the reliability.
As much as I love stories about X company being able to stuff Y capacity into storage device, the last few years have proven instructive.
1) How about doing it and producing it in such a way so that it is cheaper, not more expensive than last year?
2) How about making them at least a little bit reliable. I know you just want us to consume more and more of your drives, but lets get back to 5 year warranty's already. This one year BS is BS.
3) Maybe rather than doing the R&D to find a 60TB HD you do the R&D to find a building lot not on a fscking flood plain?
Thanks,
From everyone that bought a HD in the last year or so...
I know that there are some pretty exotic coatings in use today, but I can't help feeling, considering the number of disks likely to be sold, that this is not going to help conserve the already overstretched usage of noble metals (e.g catalytic converters etc). If Seagate bring this to market at a competitive price, then that will would be another reason to invest in noble metal mining shares (or even metal, if one can stomach the ride).
This will probably get modded all to shit since i dont have a referance, but do you think this technology will actually hit the market?
I remember back in the good old days of CD's, there was a company that created a new type of recording information to CD's using fluorescent lighting. They had a working model they presented at a technology expo, and it was all the rave saying how it will blow CD's and DVD's out of the water with the amount of storage capacity (they were able to get close to a hundred layers on a disc the size of a CD). I can't remember what the exact storage capacity was, and a google search comes up with nothing anymore. I just remember that the company suddenly disappeared (probably bought out and the tech scrapped).
Makes me wonder if the same thing might happen again, although, with a company as well known as seagate, I think it might be harder for them to disappear.
but do you ask for a 8.89 or 6.35 cm harddrive?
Also note that drive bays were named after the size of the disks that went in the drives that went in the bays. Not after the size of the bays themselves.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Only if the drive manufacturer helps you and _transparently_ does the redundancy for you at the level of reliability you want.
Because for single drive systems if you hit many bad sectors even though you have redundant copies on other sectors your throughput is going to drop to impractical rates (assuming you actually care about storing and retrieving TBs of data). The error recovery timeouts are usually in the order of _seconds_.
And if you're going to have multiple redundant drives, capacity might no longer be as important, performance and reliability might be more important. Because if your drives are too unreliable for their size you're going to need cleverer RAID controller software that won't "offline" a drive just because one sector is faulty especially on reads (it tries to read the data from all the array drives and as soon as it has enough data to build an answer, it sends it to the computer - even if a drive is not responding). Is such RAID software available?
[1] http://en.wikipedia.org/wiki/Error_recovery_control
You spread your data chunks across whole clusters of storage systems (i.e. Nimbus.io, Amazon's S3). The important data you really than need are the SHA hashes of the data.
Sure, you lose a ton of data on a drive here or there. You immediately invalidate the drive and the data on it, replication has already brought the number of good chunks of data back to the minimum replica requirement (because, you're smart, and you're keeping 3-5 copies of the same data across your storage platform), and you stream new data to the drive.
First of all, forgive my terminology: I'm quite ignorant on the subject, I just have memories of stuff read over the years.
That said, I guess the lasers included in the Seagate drive are diodes, as I recall they're just better over tubes on the cost side.
But I also remember reading that diodes have a sort of "wear" problem: over time they tend to emit less intense light to the point it gets insufficent to perform the task.
That's one of the reasons why CD/DVD drives "stop working" after a few years: the diode becomes unable to heat the disc's pits enough to perform writings first, and reads later on.
Could this impact those drives aswell? And to what extent?
Mastering the English language is fucking easy: all you have to do is to put an f* word in every fucking sentence.
I think the GP has a point, the article switches from metrical to imperial units all the time. I don't know what is the imperial unit for a nm though
Sure. Once you've picked up all of the older classic bits of content from Frys or Walmart for a song, then you can pretty much turn your back on the MPAA.
I have so much stuff that I tend to forget the stuff that I have rented via Netflix. Never mind the cinema.
Why watch the remake when the original is available and cheaper than one trip to the movies?
A Pirate and a Puritan look the same on a balance sheet.
What are the time implications of running FSCK or CHKDISK on 20TB NTFS or EXT4 ?
Duplicate and Automate.
There's really no great bother to it past setting it up. You can even get suitable and suitably simple software for this task with just about any consumer external disk.
Apple does it. Seagate does it. Western Digital does it.
There's really no good "but it's too hard" excuse here.
If anything, it's the "time is money" mentality that supports the idea of having a robust backup plan. Nothing is going to waste time more than getting caught with your pants down when you have a drive failure.
Although in that case you are likely to lose things that no amount of time or money can replace.
A Pirate and a Puritan look the same on a balance sheet.
A *really* crappy OS that is also horribly misconfigured most likely.
Also, you might want to look into setting windows explorer to run as a separate process:
http://www.addictivetips.com/windows-tips/run-windows-explorer-folders-in-a-separate-process-to-prevent-crashes-and-system-freeze/
Good luck!
people will do anything for porn.
Sweet now we can have disks with fricking lasers on their heads.
Time to offend someone
In my mind not at all. I have 2 external 1 TB drives for my music in FLAC. This is more than enough and in my desktop 2x500GB drives of which I run some very hungry programs and I can tell you that is more than plenty for me. SSD drives are good and that is the way I am going in the next few weeks. Therefore does more really mean more, or is that more headaches when it comes around to data recovery?
All cows eat grass!
I do my work from home some days with a VNC connection to work. It runs about 1.2Mbps when I'm busy (6-meg DSL).
I see almost no reason why a home user should want to have a local hard drive, except perhaps to cache media files until the upload is done (in the background, and seamless working through the cache until the upload is done, of course).
Give it a couple years and Google will offer free computers with free Internet connections in exchange for usage tracking. 70% of the population will take them up on that deal. Unless Amazon gets there first.
All that said, there's going to be a huge need for storage on the backend.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
If all else fails, use a smaller HAMR?
I'm so confused. I've been doing it wrong all these years, using a BFH to compress all those bits!
The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
I still think the answer to both the SSD and Mechanical question is hybrid drives. Seagate has tried them in the past, but they definitely aren't as fast as normal SSDs. If they can improve that tech and attach it to something like this, it's literally the best of both world. Honestly it would just be a much improved drive cache, which Seagate and other drive makers could've improved for years... but somehow never did...
will undoubtedly proove so reliable that disk drive warranties are also shrunk down to at most 1 year and eventually the target 90 days to be in compliance with all other consumer electronics.
Can you describe what a centimeter is, relative to something else?
How's this - it's the length of one edge of a cube that contains one gram of water.
At what temperature?
Every end has half a stick.
Given that transfer rates will increase by the square root of density, a 20x increase in capacity (to 60TB) will only have at most have a 5x increase in transfer rates.
Assuming that the transfer rate makes it to 500MB/s this will take 33 hours to read the contents of the drive - so something like a RAID scrub, or an offsite backup will be almost impractical.
If 10gig etherent doesn't make it to the desktop, a network backup at 1Gb/s will take a week's network bandwidth...
And even RAID will be an issue - with the loner rebuild times even with hot spares chance of a double disk failure is 5x higher than they are now (and they do happen now!)
Even if you have it backed up, reconstructing that drive would take days.
Then you are doing it wrong because in all of the backup technologies I use, I would either simply clone the backup drive (hours at most) or restore from a backup like Time Machine, also taking just a few hours...
I mean, if you were still doing incremental backup to tape I suppose. But disks to do direct clones of are cheap and then at most you are losing a week or two of data, not 60TB.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I don't know what is the imperial unit for a nm though
Glancing through Wikipedia tables, it looks like a decent candidate for that would be the size of the diamond crystal cell - 356.68pm, a nice non-round number that's almost but not quite one third of the nearest SI unit, in the best traditions of the Imperial system. We could also call them "dice" for some extra confusion.
You store your DVD collection on top of your mattress?
Like all good IT personnel he does everything in a rack.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
For comparison, it's been 60 years since IBM invented disk storage.
Its RAMAC system held 5 million characters in a stack of huge disks.
An 11-minute video of its San Jose site and RAMAC:
http://www.youtube.com/watch?feature=player_embedded&v=USJGui9yIuA
Assuming constant RPM
Why would you assume that, or the same platter size?
When we actually reach 60TB I'm pretty sure both factors will be improved.
You also totally ignored how ONE of the approaches I use is essentially a mirror. It doesn't matter if creating a new backup drive does take days, when you switch to an existing backup drive as the first step and are working in moments. If you have one local and one offsite drive there is no risk in doing that.
"There is more worth loving than we have strength to love." - Brian Jay Stanley