Stories
Slash Boxes
Comments
Search

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Beyond Nobel, Hard Drives Get Smart

Posted by Zonk on Wed Oct 17, 2007 04:34 PM
from the paging-mr.-adams-mr.-adams-to-the-security-shoe dept.
Data Storage Hardware Science
mattnyc99 writes "Giant magnetoresistance got its day in the sun when it won the Nobel Prize in physics last week—and when Hitachi rode that spotlight by announcing they'd have a 4-terabyte desktop hard drive by 2011. It's about time says Glenn Derene over at Popular Mechanics, in what amounts to an ode to the rise and future of super hard drive capacity. From his great accompanying interview with data storage visionary and computer science legend Mark Kryder: 'To get to 10 Tbits per square inch will require a drastic change in recording technology ... Hitachi, Seagate, Western Digital and Samsung ... are currently working on this 10-terabits-per-square-inch goal, which would enable a 40-terabyte hard drive.'"
+ -
story

Related Stories

[+] Science: 2007 Physics Nobel Prize For Giant Magnetoresistance 111 comments
A number of readers made sure we are aware that the 2007 Nobel Prize in physics has been awarded to Albert Fert and Peter Grunberg for simultaneously and independently discovering giant magnetoresistance. This property has allowed the explosion of disk-space growth and is cited as being one of the first nanotechnology breakthroughs. From the announcement: "Very weak magnetic changes give rise to major differences in electrical resistance in a GMR system. A system of this kind is the perfect tool for reading data from hard disks when information registered magnetically has to be converted to electric current."
Submission: Beyond Nobel, Hard Drives Get Smart (to 40TB?) by mattnyc99 (1008511)
This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.

Beyond Nobel, Hard Drives Get Smart 50 Comments More /

 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • I'd like to know which factors have allowed (forced?) the disk storage industry to continue to advance at such a steady pace. I am well aware of Moore's Law and Kryder's Law [wikipedia.org], but these are just observations, not explanations.

    Why haven't we seen similar improvements in fuel efficiency or internet bandwidth (in the US at least)?

    • Re:Steady March of Progress (Score:4, Insightful)

      by Beyond_GoodandEvil (769135) on Wednesday October 17 2007, @04:39PM (#21016775) Homepage
      I'd like to know which factors have allowed (forced?) the disk storage industry to continue to advance at such a steady pace.
      Easy pr0n, somebody should calculate how much disk space is required given mpeg2 compression to ensure that someone would have the equivalent of 60+ years of pr0n, that is how big hard disks will get.

    • Re: (Score:3, Insightful)

      by realmolo (574068)
      Yeah! And why haven't they cured cancer yet? And why does it still cost $9 for a small soda at the movie theater? Lazy-ass researchers.

      The reason that fuel efficiency and internet bandwidth haven't "increased" as much as hard drive space is because they are COMPLETELY DIFFERENT PROBLEMS with COMPLETELY DIFFERENT solutions.
    • Fuel efficiency: We have made strides forward, but American consumers seem to prefer using the added efficiency to improve acceleration rather than gas mileage. (For example, I've heard that even a 2007 Civic has significantly better acceleration and handling than the powerful muscle cars from the 50s.)

      Internet bandwidth: Huh? Ten years ago, almost everyone outside of a university was on dialup, if they had internet access at all; now, over 90% of residences have access to some kind of broadband. Sounds lik
    • Here's my stab at it, though I'm not really an expert:
      • fuel efficiency: Not knowing too much about it, I would assume that part of the problem is actual energy requirements. In order to move an object from point A to point B, you're going to need a certain amount of energy no matter what. So, if there's X amount of energy in 1 gallon of gasoline, and it takes X amount of energy to move a car Y miles, than 1 gallon of gasoline will never move a car more than Y miles. So right now, we're getting (Y-Z) mil
      • You are also not taking into account that current internal combustion engines operate at about 25-50 percent efficiency when converting the stored energy in gasoline into kinetic energy in the pistons. This is where the limited gains in mpg have come from in the last 50 years. They have improved the conversion efficiency. Who's to say that they couldn't improve it more?
      • In order to move an object from point A to point B, both placed on equal height, you need no energy at all (or, well, you need to borrow some, but you can pay most of it back when you're done). Evacuated tunnels are maybe not a realistic option, but this shows that aerodynamics and surface contact is everything. And then we haven't even started discussing the actual engine.

        • Well, yeah, sort of. My answer isn't exactly perfect, but there's something to it. Energy=work. Work can't be done without some sort of energy being done. Therefore, an object cannot be moved from point A to point B without expending energy.

          So, yeah, I guess theoretically the amount of energy needed to move an object depends on how massive it is and how fast you want to move it. Which means, in that sense, the distance doesn't matter. Right? But it takes energy to move it there at 50 MPH.

          But then,

    • Why haven't we seen similar improvements in fuel efficiency

      With regard to vehicle fuel efficiency there are other considerations in a practical everyday vehicle other than fuel efficiency. It has long been known for example that some very complex concept cars, when maintained meticulously by teams of engineers and employing technologies which are either extremely expensive, high maintenance, or impractical or all of the above, have achieved very high fuel efficiencies on the order of 70+ miles per gallo

      • Hybrid drives (Score:5, Insightful)

        by nojayuk (567177) on Wednesday October 17 2007, @04:58PM (#21017035)

        I believe the higher capacity drives will force a rethink on how data is stored and accessed on standalone machines like laptops and desktops. I've only got a couple of terabytes of data on this machine and doing a file search over the five (I think, I can't actually remember how many drives I've got fitted in this thing) disks is already pretty time-consuming. The solution will be to add intelligence to the disk interface so that data indexing is done pre-emptively and the results cached on the fly.

        The first generation of hybrid drives are already here but they're only at the beginning of their development cycle. HDD recording densities will increase as will flash RAM densities and that will improve access times but only for the most commonly accessed data.

        Imagine a 10Tb HDD built in the classic 3.5" wide form factor, with 256Gb of 1024-bit-wide 150MWord/sec flash memory or MRAM on the controller board acting as cache. The spinning disk becomes a backing store for the flash where data is kept "fresh" by a smart algorithm. The drive spins down intelligently when not needed, saving power and reducing heat dissipation.

        Higher recording densities are only one part of the future of disk drive technology.

        • Re:Hybrid drives (Score:4, Interesting)

          by ArcherB (796902) * on Wednesday October 17 2007, @05:51PM (#21017773) Journal
          Imagine a 10Tb HDD built in the classic 3.5" wide form factor, with 256Gb of 1024-bit-wide 150MWord/sec flash memory or MRAM on the controller board acting as cache. The spinning disk becomes a backing store for the flash where data is kept "fresh" by a smart algorithm. The drive spins down intelligently when not needed, saving power and reducing heat dissipation.

          I'd rather they be broken into separate drives. I'd like a flash based drive for my OS and maybe a few commonly used applications and a spinning HDD for all my data and backups.

          • Re: (Score:3, Informative)

            by Jake73 (306340)
            You know, this is a pretty interesting point.

            Perhaps it would be better to have two drives available to the OS with rated latencies and bandwidths. Then, the OS can make software-based decisions based on the usage profile of the machine (server, workstation, media, etc).

            Alternatively, some rating could be given to each file installed by software installation programs. Things like help databases, samples, aux tools, uninstallers, etc could be thrown on lower-latency spin disks. The critical items like pro

      • Progress in new directions (Score:4, Interesting)

        by mollog (841386) on Wednesday October 17 2007, @05:19PM (#21017301)
        When I see headlines about 1TB drives, I immediately think of losing 1TB of data.

        How about they put a RAID 1 array in a 3.5" form factor? Two separate platters, two head/arm assemblies, two SATA connectors.
        • Because you could just buy two 3.5" drives and run them in RAID1 yourself? It only costs money.

            • Re: (Score:3, Insightful)

              by ckedge (192996)
              Imagine running a raid-1 array where if one half of the array goes bad, you have to replace BOTH. This immediately doubles your failure cost. No way in heck is anyone going to do that. Instead their going to raid-1 their failed "integrated raid-1" drives until the second half fails itself. Anything else would be a gross waste of money and time.

              > You have to do that yourself,

              How is buying and plugging in a second drive "hard"? You already have to buy one, why not just tell the guy behind the counter
      • 7200 RPM gives you a hard wall. Faster rotation is a pain, and arm movement is not (generally) the limiting factor. The only thing I can imagine is putting two heads there, right opposite each other. That creates a nice scheduling problem, but I guess it would be doable. You wouldn't only get RAID 0, because, with two heads free, you could actually cut the time before the right sector is under either head in half. One thing that comes to mind is whether a construction with two arms would be much more (i.e.
        • WREN did this in their WRENiii series ESDI drives. For the time those drives were fast. Rather than a radial axis for the drive heads they used a linear actuator where the head stacks could move independently of each other. Worked like a champ. I would still use them but the controller is VLB (and no MB these days supports that on P4's and up), and the drives were only 160 meg (and 5-1/4" full height beasts).

          -nB
  • that video cards will get better, games will get much larger, and of course - we'll all be fighting over which format the game will be on and complain if its blue-ray and we have HD-DVD, and vise versa.
    • No. HD-DVD/Blueray will likely be the last generation of physical media.

      Oh, there'll be media offcourse, you need to store stuff -somewhere-, but bundling the data, which is what you pay for, and the storage-medium is no longer interesting, makes about as much sense as selling water, and insist it only be stored in YELLOW bottles, not BLUE ones damnit.

      Music, Movies, Software, these are all just data. Where I want to -store- my data is up to me, I will choose based on price/performance/convenience, but my ch

    • we'll all be fighting over which format the game will be on and complain if its blue-ray and we have HD-DVD, and vise versa.

      Complain ?
      Why complain ?
      By then, most users and all /. will have no-name korean multi-format burner that will handle both HDDVD and BlueRay.

      HDDVD and BlueRay is no real format war and has nothing to do with the old VHS vs. BetaMax stuff. It's closer to the DVD "plus" vs. "minus", because with disc, multi-format are easily doable.

      And are actually already done, several companies have ano

  • 2-Way Wrist HD (Score:4, Interesting)

    by Doc Ruby (173196) on Wednesday October 17 2007, @04:43PM (#21016829) Homepage Journal

    this 10-terabits-per-square-inch goal, which would enable a 40-terabyte hard drive.'

    It could also enable a 750-gigabyte 1" radius HD, if they're really clever. Which could serve the Bluetooth wristphone/player we've all been waiting for. So we can stop referring to that mobile multimedia terminal as a "phone", and again more accurately as a "watch".
    • It's sad to think we're going to need drives this size just to install Ubuntu in 2011.

      And you thought I was going to say Windows!

  • For the average person (Score:3, Insightful)

    by Tribbin (565963) on Wednesday October 17 2007, @04:51PM (#21016931) Homepage
    The biggest part of our hard disks are spent on movies, music and games.

    Most of these are on thousands of computers.

    Wouldn't a good sharing/streaming protocol/project be the solution for storage for the average person?

    • Re: (Score:3, Insightful)

      by Zironic (1112127)
      For now and probably for quite a while disk space is cheaper then bandwidth.

    • Of course, the problem with that is the idea of "copyright". If we were most concerned with conservation of resources, reliability of backups, and easy distribution, we probably would have made a huge shared filesystem using methods similar to bittorrent, and all movies/music would be stored online and made readily available to anyone with an internet connection. Storing this stuff on your local hard drive would be only necessary for the purpose of caching it so you could listen offline.

      Still, big hard d

  • 40-terabyte hard drive (Score:4, Funny)

    by Nonillion (266505) on Wednesday October 17 2007, @04:55PM (#21016999)
    Yeah, in 10 years well be bitching because we wont have enough space for a decent Windows xx or Linux xx install. I must be getting old, because I remember asking how on earth could you fill a 40 Megabyte hard drive.

    • It was always easy if you were into audio recording. Digital video has, of course, compounded this a lot, but I've wished I could increase the size of my storage for about as long as I can remember. Even for plain old text, 40 megabytes isn't a lot. Are you telling me you didn't have more than 50 3.5" floppies?

  • this filesystem has been mounted 32 time, checking filesystem.
    634 Hours Remaining.
    • There are plenty filesystems that don't do that, which are included with any kernel released in the last few years.
      • I'm not sure I understand.

        As I know, you want the ability to check your file system in case of a dirty shutdown. For example, if you still have data in RAM pending write-backs to the drive and have a power failure; it would be a good idea to check the partition for data corruption. Luckily most modern file systems are "journaling" which helps prevent against corruption. Either the data has been committed, or it hasn't.

        • Luckily most modern file systems are "journaling" which helps prevent against corruption. Either the data has been committed, or it hasn't.

          My point exactly. Any kernel made in the last few years supports a journalling filesystem which recovers in seconds from a dirty shutdown. It probably even recovers faster on larger disks as the journal is usually of a fixed size, and bigger disks are usually faster than smaller ones.

          The grandparent is talking about what would happen if you used ext2 for a drive with a s

  • How do you back it up? (Score:3, Interesting)

    by HonkyLips (654494) on Wednesday October 17 2007, @05:13PM (#21017223)
    I work in animation & video production and a single project can take up a terabyte... I'm all for storage increases but I have no idea how to back it all up... It's all very well for the Blu-Ray and HD-DVD club to go on about storing 30/50 gig on a disc but when your drive holds 4 terabytes (and you just know it will fill up quickly) the backup problems just get bigger too...
    • Backing up is essentially straightforward....use more hard drives.

      The real problem is that the transfer rate is not keeping up with the capacity increases, so the amount of time it takes to fully duplicate a drive keeps going up. Maybe it's time for multiple heads per platter, kind of like the 72X CDROM drive from a while back.
    • mirror.

      High performance tape library.

      Keep peoples works local and on a central server.

      There are solutions. I get 550 an hour to consult, let me know if you need any work done.

  • Why Not Even Bigger? (Score:4, Interesting)

    by Ralph Spoilsport (673134) on Wednesday October 17 2007, @05:24PM (#21017383) Journal
    If you can get a TB on 1 inch drive, why not build a drive with the same density that's LOTS bigger? I imagine warping might be a problem, but I remember 10" winchester drives!

    A 2.4cm drive has an area (just for this thought experiment) of (1.2 x 1.2)pi, or roughly,4.52 sq cm. now, a 10 inch drive (24cm) has an area of 45.2 sq cm.

    So, that would make it a 45 TB drive. Data retrieval might be kind of slow, but: if you have massive RAM caching, it could be of great use. Imagine a home theatre with something like this.

    Imagine buying a drive like this that comes pre-installed with every song ever produced by WEA or EMI or Sony/Columbia. Say, everything from 1925 onward. How much would you pay for such a drive?

    Or, ALL the movies ever made by (name your favourite) movie studio between (date x) and (date y).

    I'd pay some serious green for that. All the classic movies. All the great songs of history.

    That's what we're facing, very very soon: the trivialisation of media technology.

    And eventually, that 25cm drive holding 45TB becomes a 2 inch drive holding 90TB.

    We should be able to predict the arrival of the $500 2 inch exabyte drive.

    The entire collection of world culture, audio in mp3, film in mp4, and images in jpg. Japanese, chinese, American, canadian, English, French, Italian, Russian, etc etc etc. on one or maybe two drives, or even one for audio, one for video, and one for images.

    what then? with all of audio and visual culture at your fingertips, what will we do with it? what will a society in the future (assuming it doesn't implode with the loss of petroleum, or vapourise itself fighting over it) DO with that much data commonly available. to anyone?

    Will it be possible to write a new melody? Will it be possible to tell a new story? Will it be possible to make an image that matters? Some would argue that imaging is dead - eaten alive by advertising. some would argue that film is dead as all the stories are told, and now we're in a grid of "1 from column A, two from column B" kind of mix and match story telling. And some say that even music itself has run its course - washed up on the blandishments of pop, the inaccessibility of the academy, and the dumbed-down rumbling of a sold before it was born hiphop, and an inchoate melange of world music that mimics and fights the imperial culture.

    When it's ALL on your drive, who cares? will culture just gradually wither away?

    Maybe we will do better when the oil runs out, and the machines stop working. We'll have to sing to each other, and tell stories to each other by the fire, instead of the sitting around having the fire tell stories to us.

    RS

    • Re: (Score:3, Informative)

      by sconeu (64226)

      A 2.4cm drive has an area (just for this thought experiment) of (1.2 x 1.2)pi, or roughly,4.52 sq cm. now, a 10 inch drive (24cm) has an area of 45.2 sq cm.


      Math error. 452 cm^2. Remember, you're squaring that 10.
    • Yeah, true, pretty soon you'll be able to store all human culture on your wristwatch. Well, except for whatever is made in the last decade, because offcourse the bandwith of media will keep going up with storage. (a Blueray disk takes more space than a DVD. Whatever we have in 20 years for movies will take more space than blueray does)
  • The real problem that I see is that drive bandwidth has not been increasing at the same rate as drive capacity, which means that the time to read/write an entire disk keeps going up.

    Maybe it's time that manufacturers start using multiple heads per platter to cut down on seek times and increase bandwidth. I'm sure there are people that would pay for double the bandwidth...why hasn't anyone done this yet?
    • Transfer rate is limited by the bandwidth of the electronics (preamps and channel.) After all, you have to pick up microvolts, amplify them to usable amplitudes with low noise, and then do A/D sampling and some fairly complex filtering on them. Today's drives transfer at around a Gb/s; that is not going to increase much. Nobody will want to pay for or cool GaAs read channels. And there's no reason to expect that seek times can be reduced much. Latency could be reduced further in exchange for higher power co

  • Yeah, but... (Score:4, Insightful)

    by unitron (5733) on Wednesday October 17 2007, @05:56PM (#21017847) Homepage Journal
    That's great and all, but will we still be limited to 4 primary partitions?

  • Danger! (Score:3, Funny)

    by RowanS (1049078) on Wednesday October 17 2007, @07:42PM (#21019103)

    Giant magnetoresistance got its day in the sun when it won the Nobel Prize in physics last week--and when Hitachi rode that spotlight by announcing they'd have a 4-terabyte desktop hard drive by 2011.
    Oh my god! Four terabytes of sentences like that would contain over 6 x 10^10 mixed metaphors. Crammed into a single 3.5" drive bay the figurative density would be so great that the drive would collapse into a metaphorical black hole, sucking in all nearby figures of speech, similes and allusions. Somebody stop them!
    • In other words: "1TB ought to be enough for anybody"

      We know how that ended ;)
        • The average household's needs don't change that much. Can you honestly say they fill even a 160 GB HDD?

          Lots of average households fill more than that with nothing more than an HD DVR, without even getting into movies, home videos, photos or music.
    • Your not thinking like a marketer! Most households also don't not need a frigging TB of HDD space. We're providing a value-add edge over competitive products while maintaining sales figures consistent with current products for a win-win solution to all parties! We're selling the new models at the price of the old model while the old model gets cheap/fades away and the customer gets extra space.

      Now, if you'll excuse me, I need to go erode some pillars with my head (gets that awful marketing after-thought ou
    • "Most households don't need a frigging TB of HDD space."

      Just like how 640K of ram was "enough for everybody"... hmmm. Just like how supereme commander doesn't need 64-bit memory addressing... either way, the thing is we'll find ways to use it and the consumer is not the only customer of hard drive technology don't forget. The medical and scientific community need enormous amounts of storage for the volume of data that is being generated for research purposes.

      I'm already filling up over 2 terabytes of hard
    • Home movies and picture.

      • Re: (Score:2, Insightful)

        by BenoitRen (998927)

        It just seems stupid to me to buy a HDD larger than you need. They should focus more on performance and reliability instead of size.

        By the way, my non-development PC that I use most has a 6.2 GB HDD, which is currently barely 1.5 GB full.

        • Reliability and performance? Build the disk with three platters, each with independent head control on both sides. RAID in a box. You'd get the performance advantages of RAID without the inherent disadvantage of having n times as many spindles to develop bearing problems. I'd certainly buy a terabyte disk with built-in RAID 5+1. That would rock.

    • That won't help. Most hard drives fail catastrophically when they fail. If you have just a few sectors going bad, SMART will tell you before you actually try to read the data if you have background scanning turned on.

    • The hope in increasing storage capacity is that at some point you'll hit a magic number that will basically mean "unlimited" for your needs. One example is documents. It used to be in the old days you could easily fill up a floppy disk with documents so you started fiddling around with multiple floppies to store your crap. The same was true to some extent for CDs. Once people started burning CDs. Suddenly you had collections of CDs filled with audio and data because each disk simply did not have enough cap

All trademarks and copyrights on this page are owned by their respective owners. Comments are owned by the Poster. The Rest © 1997-2009 Geeknet, Inc.