Slashdot Mirror


60TB Disk Drives Could Be a Reality In 2016

CWmike writes "The maximum areal densities of hard disk drives are expected to more than double by 2016, according to IHS iSuppli. Hard drive company Seagate has also predicted a doubling of drive density, and now IHS iSuppli is confirming what the vendor community already knew. Leading the way for greater disk density will be technologies such as heat-assisted magnetic recording (HAMR), which Seagate patented in 2006. Seagate has already said it will be able to produce a 60TB 3.5-in. hard drive by 2016. Laptop drives could reach 10TB to 20TB in the same time frame, IHS iSuppli stated. It said areal densities are projected to climb to a maximum 1,800 Gbits per square inch per platter by 2016, up from 744 Gbits per square inch in 2011. Areal density equals bit density, or bits of information per inch of a track, multiplied by tracks per inch on a drive platter. This year, hard drive areal densities are estimated to reach 780Gbits per square inch per platter, and then rise to 900Gbits per square inch next year."

21 of 293 comments (clear)

  1. For depressed people by SadBob · · Score: 5, Funny

    Since pirates are depressed people, these will be perfect fit for depressed pirates.

  2. WOW by __aaeihw9960 · · Score: 4, Funny

    That's a shitload of porn.

    1. Re:WOW by Nyder · · Score: 4, Funny

      That's a shitload of porn.

      I must be a nerd, because my digital comic collection is bigger then my porn collection.

      --
      Be seeing you...
    2. Re:WOW by Anonymous Coward · · Score: 5, Funny

      I must be a nerd, because my digital comic collection is bigger then my porn collection.

      For some people, those would be one and the same.

    3. Re:WOW by xQx · · Score: 5, Insightful

      Truecrypt.

      That's all I have to say on the matter.

    4. Re:WOW by MightyYar · · Score: 5, Funny

      I use Spousal Truecrypt. It's an unencrypted folder titled "Sports".

      --
      W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
    5. Re:WOW by NaughtyNimitz · · Score: 5, Funny

      Until your smart wife discovers your loot when she types "creampie" or "cup" in spotlight (or similar windows search engine)

    6. Re:WOW by Jafafa+Hots · · Score: 4, Funny

      Don't assume like that. There are people into ANY kind of porn, whether high definition or high density.

      --
      This space available.
    7. Re:WOW by Alter_3d · · Score: 5, Funny

      I use Spousal Truecrypt. It's an unencrypted folder titled "Sports".

      A friend of mine kept his porn folder on plain view. The folder title was "Uninstall Windows". The wife never looked inside.

  3. More capacity, but what about I/O? by mlts · · Score: 5, Informative

    One thing we have had issues with is that even now, the issue with drives is how fast we can get data in and out of it.

    Even the high end SAN makers know this and tell people to always use RAID 6 on the backend, just because the window of time that it takes to rebuild a drive is so long these days that it can easily allow for a second drive failure to happen with no protection.

    What I really will dread seeing is an external 60TB drive that is stuck with a USB 3 interface as its only I/O. USB 3 (for lowest denominator compatibility), a SATA descendant, and Thunderbolt, would be ideal, but with how cheap some drives end up, it might just be a sole USB port for in/out.

    1. Re:More capacity, but what about I/O? by Sancho · · Score: 4, Insightful

      Even the high end SAN makers know this and tell people to always use RAID 6 on the backend, just because the window of time that it takes to rebuild a drive is so long these days that it can easily allow for a second drive failure to happen with no protection.

      It's not just another drive failing--it's unrecoverable read errors (UREs). You might not know that a sector is unreadable until it's too late--if you discover it during a resliver of a RAID5, you are seriously out of luck. With very high data densities per disk, the chances of a URE are high.

      So you're right--I/O speed is important. Also important is resiliency. If these don't scale along with the sizes, I think these will be considerably less useful than most people hope.

    2. Re:More capacity, but what about I/O? by Conception · · Score: 4, Interesting

      It's not the interface, it's the drives themselves. They aren't really faster than they were when ATA-133 came out. Doesn't matter what interface you stick on there, hard-drives aren't getting faster (thank god for SSD). At 60TB also, the BER rate approaches something like 600% chance over the whole of the drive, or something like that, if they are using the same reliability numbers that current drives use. Terrifying.

    3. Re:More capacity, but what about I/O? by AshtangiMan · · Score: 4, Funny

      I agree that RAID is not a backup, but that doesn't mean that my backup can't be a RAID . . .

    4. Re:More capacity, but what about I/O? by afidel · · Score: 5, Interesting

      Actually as areal density increases drives are getting faster, both it terms of streaming reads and average transfer time, it's only worst case performance that is not getting any better. Also the drive manufacturers aren't stupid, as physical density increases logical density isn't increasing as quickly because they are using a larger percentage of the physical bits for error correction meaning the logic BER should at worst remain constant.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
  4. I'm going to make a bet or three by CAIMLAS · · Score: 4, Interesting

    I'm going to make several bets here which will also hold true:

    * sequential performance will improve at a rate congruent with storage capacity
    * random performance will remain roughly the same as it has for the past 10 years (ie, poor, though it will likely improve slightly unless we go back to double-thick drives like we had 10-15 years ago)
    * resiliency will not improve for single disks and will likely be worse for in terms of longevity.
    * none of this will matter for the consumer market, because by that time, everyone will be using SSDs almost exclusively. You can still fit a lot of data on a 500GB drive, and those are commonly available for laptops and desktops already.

    --
    ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    1. Re:I'm going to make a bet or three by Spodi · · Score: 5, Insightful

      And I bought 2x 2TB HDDs for less than half that. Your point? Why do some people have such a hard time understanding that not everyone cares about speed for all of their drives. My primary drive, sure, make that baby as fast as possible. But all I need there is 200 GB (85 GB at this time) since that just holds the OS and all programs I use. The rest - the multiple TBs of backups and media (music, movies, pictures), who cares how fast that is. Even the slowest HDDs are going to be able to play 1080p just fine. For the very rare occasions those drives bottleneck, I don't mind waiting. I'd rather spend the money upgrading everything else that bottlenecks far more often.

  5. Re:What's the useful limit? by mpetch · · Score: 5, Funny

    640TB

  6. Yeah f'ing right by rgbrenner · · Score: 4, Insightful

    2016 is in 4 years. Let's see...

    In 2008, Seagate announced the world's first 1.5TB drive.

    And in 2012, Hitachi announced the first 4TB drive.

    And in 2016, this will magically become 60TB?!

    If you said 10TB, I would believe it. I'll even go along with 15TB.

    But 60TB? don't believe it for a second.

  7. Re:I don't get it. by atrain728 · · Score: 5, Informative

    The numbers the summary cites are Gbits per square inch. Meaning it's already been squared.

  8. Re:I don't get it. by MickyTheIdiot · · Score: 4, Insightful

    HELL YES! Bring back the Quantum Bigfoot!!!11!

  9. Re:I don't get it. by jsm300 · · Score: 5, Informative
    No, the article is quoting aureal density which is expressed in gigabits or terabits per square inch. The problem with the article is that it is combining data from various sources and misreading/misinterpreting the data (so what's new, this is Slashdot after all).

    First, the summary above says that Seagate will produce a 60 Tb drive by 2016. That is not true. Seagate has said they will produce a drive with "up to" 60 Tb of capacity (30-60 TB) by the end of the decade. This is based on the theoretical limits of HAMR technology, which are projected to be in the 5-10 Tbits/sq. inch. range. Current 4TB drives are made with platters that have a density of around 650 Gbits/sq. in., so the math works (10Tb/.65Tb is approximately 15x).

    The other part of the article is talking about what the maximum density is likely to be over the timeframe from now to 2016 using PMR technology and transitioning to something new like HAMR. PMR technology will top out at about 1Tbit/sq. inch, so anything over that will require something new like HAMR. that underlying article quotes 1.8 Tbit/sq. in in 2016, which may not be out of line with 5-10 Tbit/sq. in. by 2020 as a new technology like HAMR comes online.

    The two articles that I am basing the above on are:
    Seagate/HAMR article
    IHS/ISuppli article