Please create an account to participate in the Slashdot moderation system


Forgot your password?
Data Storage Hardware Technology

60TB Disk Drives Could Be a Reality In 2016 293

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."
This discussion has been archived. No new comments can be posted.

60TB Disk Drives Could Be a Reality In 2016

Comments Filter:
  • by SadBob ( 2645421 ) on Tuesday May 22, 2012 @05:45PM (#40080931)
    Since pirates are depressed people [], these will be perfect fit for depressed pirates.
  • WOW (Score:5, Funny)

    by __aaeihw9960 ( 2531696 ) on Tuesday May 22, 2012 @05:46PM (#40080941)
    That's a shitload of porn.
  • I don't get it. (Score:2, Interesting)

    by Anonymous Coward

    If 4TB is the biggest drive you can get today, wouldn't densities have to increase by 15x to get to 60TB drives by 2016, not just "more than double"

    • Re: (Score:3, Informative)

      by dgatwood ( 11270 )

      No, you'e thinking too linearly. The density increases in two dimensions, so the capacity increases by the square of the density (approximately). You would need just shy of a 4x increase in capacity without increasing the number of platters. If you can find a way to decrease the spacing between platters, you could get a 15x capacity increase with an even smaller density increase.

      • Re:I don't get it. (Score:5, Informative)

        by atrain728 ( 1835698 ) on Tuesday May 22, 2012 @06:18PM (#40081313)
        The numbers the summary cites are Gbits per square inch. Meaning it's already been squared.
      • Re:I don't get it. (Score:5, Informative)

        by jsm300 ( 669719 ) on Tuesday May 22, 2012 @07:29PM (#40081893)
        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 []

    • If 4TB is the biggest drive you can get today, wouldn't densities have to increase by 15x to get to 60TB drives by 2016, not just "more than double"

      Probably. The first 3tb was released June 2010. [] 4tb came out Oct 2011. [] Not exactly amazing growth, over a year for 1tb, at this rate we'll be 9tb in 2016. At this rate we will not see 60tb by 2016, and I say "we" meaning end consumer, maybe some lab monkey will see an areal density equivalent to 60tb, but it won't be available for sale. And for anyone wondering the answer is yes, the 4tb drives already use five platters, 800gb each, [] so they can't shove more platters in there to double capacity currentl

  • by mlts ( 1038732 ) on Tuesday May 22, 2012 @05:47PM (#40080963)

    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.

    • Re: (Score:2, Funny)

      by jedidiah ( 1196 )

      No. The dreadful part is not being stuck with USB3.

      The dreadful part is realizing that attaching an over hyped external interface to it will likely not matter.

    • by Sancho ( 17056 ) * on Tuesday May 22, 2012 @05:54PM (#40081015) Homepage

      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.

      • Agreed. I recently was reading that in a Raid 5 array with 2TB disks it is likely that you will lose data in the event of a single disk failure because of UREs. I have been thinking that setting up various raid 10 spindles is the best way to archive and protect data. One spindle for use, and another for backup, all with 1TB drives (2TB of space with 4 1TB drives). It strikes me that HDD is a terribly inefficient way to back up data securely, but surely it is better than optical disks.

        • by Sancho ( 17056 ) *

          The old wisdom is that "RAID is not a backup." So RAID5, RAID6..doesn't matter, you should be backing up regardless.

          Tape is still probably the cheapest medium for backups, though you have to invest in the drives as well (and those can be in the thousands of dollars.)

          I use online backups to a remote host along with a large drive which is offline most of the time and rotated with a few others. I'm just getting into ZFS and snapshots, with which I can more easily take frequent backups. Unfortunately, I don't

      • by Hatta ( 162192 )

        It's not just another drive failing--it's unrecoverable read errors (UREs).

        A URE is a type of drive failure.

        • by Sancho ( 17056 ) *

          It's not what most people mean when they say, "Drive failure," and the URE could have happened before the RAID was ever put into degraded mode (it could have been the first failure, just no one noticed it.)

          • Which is why it's very important to monitor your disks using the tools and the SMART data on the disks themselves.

      • by mcrbids ( 148650 )

        I've been hearing about the end of hard disks for a long, LONG time now. RAID1 was supposed to be dead a decade ago, RAID5 a few years back.

        Strangely, rumors of their deaths have been repeatedly found to be greatly exaggerated. Speeds have improved, (though not kept pace with the sizes of drives), and so has reliability-per-bit.

      • by DigiShaman ( 671371 ) on Tuesday May 22, 2012 @07:11PM (#40081789) Homepage

        At least with Dell PERC controllers, the likelihood of UREs these days have been mitigated with the use of background patrol reading which proactively checks the disk in idle periods or low priority disk access. UREs are also detected and corrected on the fly while accessing data.

        You might want to check the documentation as to what features your RAID controller supports. If it supports background patrol reading, most likely it's enabled by default. If not, I suggest upgrading the controller if possible.

    • by Conception ( 212279 ) on Tuesday May 22, 2012 @06:00PM (#40081083)

      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.

      • by afidel ( 530433 ) on Tuesday May 22, 2012 @06:59PM (#40081687)
        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.
    • by Tynin ( 634655 )
      You just have to make your platform able to handle file replications to N + 1 nodes (or any of the other plethora of ways you can slice it). Then go with a RAID 5 or whatever strategy you want, to give individual nodes the likelihood of higher availability. If you need to handle higher IOPS you can go with a distributed filesystem that can handle metadata/journalling on discrete nodes and load it up with SSD's / Fusion IO, whatever fits your bill. Then you can have banks of slow disks without much concern o
    • It's not as bad as it may seem. With disk speeds up to 15,000 rpm and higher areal densities means that data can be pulled off pretty fast. If HDD manufacturers were to implement technologies such as multi-track disk heads then IO speed could increase a lot more and would be limited mainly by seek times. What a lot of companies are doing nowadays is using 2" (laptop) drives in their servers, packing a lot more drives into the space, which means more smaller disks and therefore less to rebuild in the even

    • RAID 10 is better than RAID 6.
      Rebuild time is nothing, your performance doesn't degrade to shit, and you're more than likely going to survive a second failure during rebuilding.

      If you're paranoid you can run multiple mirrors.

      • by afidel ( 530433 ) on Tuesday May 22, 2012 @07:08PM (#40081761)
        Without parity you're going to miss certain types of corruption so RAID6 is actually superior from a data reliability standpoint.
        • Finally got my new NAS box up this past weekend... 12x 3TB drives.. 10 of them in RAID-Z2 (FreeNAS) with the other 2 has hot spares. 22.5TB of raw storage, and hopefully should be enough space for the next 4-5 years.. I've been aquiring/using about 1-2TB/year for the past few years. Though 60TB drives would be cool.. the replication time would suck.
  • by neokushan ( 932374 ) on Tuesday May 22, 2012 @05:50PM (#40080983)

    Ok, there's never going to be a hard drive big enough to suit everyone's needs - that's a given. But average joe consumer must have a limit of some kind - what is it?
    I can't see how an average person will use more than about 1TB of space any time soon and even then that's probably overkill. At one point maybe it would have been to store music and films, but that's going to the cloud rather than local storage. Average joe doesn't rip his blu-rays.
    In the same way that RAM has probably hit a peak with consumers who simply don't need more than 3 or 4Gb for what they want to do, I wonder how Hard drives will fare?

    Now as for myself, I could definitely fill 60Tb of space with stuff I'd like to keep - sign me up, but with the price of SSD's seemingly halving over the last couple of months, it's only a matter of time before average joe customer starts to realise that for the same price of a 60Tb HDD, they could probably have a 1Tb SSD that's a lot faster.

    • Re: (Score:3, Insightful)

      HD video, music, photos. ETC. Even grandma can fill a 1TB hard drive with HD video without even trying.
    • by jedidiah ( 1196 )

      If "Average Joe" has any awareness of speed, then the Cloud will quickly get kicked to the curb and greater local storage densities will matter.

      "Average Joe" will likely never realize that there is a technical reason to seek out an SSD. Some marketing hype might push them in that direction. Genuine "geeky" technical understanding will not.

      Joe is willing to tolerate the cloud but wants the speed of an SSD? That's a clear an obvious contradiction.

    • by vlm ( 69642 ) on Tuesday May 22, 2012 @05:58PM (#40081067)

      I can't see how an average person will use more than about 1TB of space any time soon and even then that's probably overkill.

      video editing. 1TB is about one of my wife's typical projects. What the "creative" types don't realize is if you record 10,20,30 times as much "stuff" as makes it into the final product, to edit you've got to store all that junk somewhere.

      There are batching strategies where you can edit a three hour long interview down to 5 minutes of actual usable clips, repeat until everything is "clipped", then merge up all the clips and edit those. Some video editing software is very unhappy with terabyte scale projects so you have to do this anyway.

      You can't edit and dispose of interview #4 because someone might have a cool story to run against it in interview #35.

      This is not crazy stuff either, family history stuff

      • Well of course, but you're one of the many who could make use of such space - I'm talking about average only-does-email-and-facebook joe kind of person.

      • Yea, I make a few movies and I'll have to record like 40 min for about 5-10 min of final footage.

    • Average joe doesn't rip his blu-rays.

      Mostly, because it takes up too much room on a drive, and too slow to download from the internet :)

    • by 6ULDV8 ( 226100 )

      I've found that a larger hard drive just increases the likelihood of me storing redundant data on the same media. I get lazy about housekeeping. I recently pulled the 500G drive from my laptop and replaced it with a 120G SSD. Instead of carrying everything I own, I only carry tools I need. Now I can actually manage a daily backup to a NAS and not have to wait while it completes. If I'd had the extra cash, I would have likely purchased a larger SSD and still be carrying all the cruft that I haven't touc

      • by jedidiah ( 1196 )

        This is just sad. You can't even carry your music collection with you and you are trying to make up sad and pathetic rationalizations as to why it's actually a good thing.

        Tech should adapt to you rather than the other way around.

        • Tech should adapt to you rather than the other way around.

          The opposite is actually working pretty well for Apple these days. You just have to market it correctly.

        • Personally, I've got my entire music collection available to me via Google Music. It took me ages to upload the 18k or so tracks (a week of overnights with upload maxed out), but now that it's all there, I've got it on anything running a web browser as well as my phone.

          Now, I would never in a million years dump the files stored locally, but I sure as shit don't bother loading GBs of music onto an MP3 player or my phone when I go on trips anymore, nor waste time burning CD's for the car or any of that crap.

      • Precisely. When my dh0: drive got full, I thought about copying it over to usb0: but when I looked at what was there, I realized I didn't need "Beauty and the Geek" or "Transformers 2" or "Billboard's Hot 1000 songs of the 1960s" so I erased them. I freed-up about 100 GB of junk I never should have kept in the first place.

      • Your music collection fits on a 500G? Your aren't even trying.

    • They can have both. And will.
    • by Korin43 ( 881732 )

      Hopefully this will also lower $/GB and decrease cloud storage prices as well. That's what I'm excited about.

    • by Hentes ( 2461350 )

      It varies from person to person. I have 700GB storage and don't even use all of it.

    • by rsborg ( 111459 ) on Tuesday May 22, 2012 @06:09PM (#40081197) Homepage

      Ok, there's never going to be a hard drive big enough to suit everyone's needs - that's a given. But average joe consumer must have a limit of some kind - what is it?

      Thing is, there are multiple "average joe users". Just from my knowledge I could state about 4-6 profiles which have different processing, portability, storage and interface needs. My dad is chugging along fine with his MB Air, but despite that sweet chassis, I need more local storage and more RAM.

      To apocryphally quote a famous person, 64.0GB is enough for most people... and I'm sure both you and I are not "most people".

    • I've noticed that when friends get faster internet access (100/10 or even 100/100) they tend to download less and less, and just stream the shows/movies they want to watch. For them loads of storage is not really necessary.
      Of course you still have the odd one who wants to download everything in 1080p x/h264 rips at 8-20 GB a piece.

      So, for people with slower speeds I can see they'd want to have more storage, as streaming is not an option. (no one rents discs anymore, there's a reason the shops are closing)


      • While I download less and less, it is because I archive everything and so I have less stuff to download. I like 1080p rips, but do not hunt for them, 720p or even SD can be good enough (though I will download the higher resolution version if I get the option to).

        Streaming depends on other people, the data I have only depends on me. I may choose to stream instead of grabbing the tape, but I want the tape to be there, in case I cannot stream (internet connection goes down, the streaming provider goes down or

    • by mlts ( 1038732 ) *

      If a drive has capacity enough to back up completely at least one of my machines, that would be considered adequate by me. Having the 1TB SSD is nice, but so is having a pair of drives that data is copied to on a nightly basis so if the SSD gets lost or erased, it is still present (without having to waste the time and bandwidth costs to recover from a cloud provider.)

      Maybe we will see disks with more options for interfaces. For example, it would be nice if an external drive could be configured to show up

    • I can't store 1TB of movies and music on the cloud, this would be a very high recurring fee and I would upload data at 120KB/s. a one-time fee for the HDD and another one for the backup HDD is cheaper. I still get to access files from my cheap ass NAS with VIA CPU, even though at 120KB/s from outside but when I get fiber it will be more like 10MB/s.

      I do want to get a 4TB HDD for it, RAID is not so good as you still have to buy one or more backup drives so there aren't any savings. maybe put the OS on a SD c

    • >>>Average joe doesn't rip his blu-rays.

      No but they might download a pirate rip of a BRD. :-)

      >>>RAM has probably hit a peak with consumers who simply don't need more than 3 or 4Gb for what they want to do

      It's not the average Joe who is carelessly consuming RAM. It's the programmers. I remember when I bought my PC in 2002 and it had half-a-gig of space. That was almost 10 times more than the minimum recommended by Microsoft XP. It ran superfast! But NOW the Flash has grown, the browse

    • >>>I can't see how an average person will use more than about 1TB of space any time soon

      P.S. If I wanted to be really anal, I could argue that my Commodore 64 played music with only 0.00006 GB of memory, and my Amiga did videos with only 0.00025 GB of memory. In consoles, Super Nintendo and Sega Genesis only used ~.004 gigabyte cartridges. Why on earth would anybody need more than that? Answer: Media grows in size and needs more storage space.

  • I've been waiting for this day.
  • by CAIMLAS ( 41445 ) on Tuesday May 22, 2012 @05:54PM (#40081019) Homepage

    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.

    • by jedidiah ( 1196 )

      A 500G SSD costs more than the laptop or desktop you would want to connect it to.

      You seem to be pining for something that you have no real awareness of.

      • A 500G SSD costs more than the laptop or desktop you would want to connect it to.

        You seem to be pining for something that you have no real awareness of.

        Wrong. About a year ago I bought two high-end 256 GB SSDs for less than $500.

        • by Spodi ( 2259976 ) on Tuesday May 22, 2012 @10:29PM (#40082809)
          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.
      • 512GB SSD for $500 []

        • by fnj ( 64210 )

          Like the guy said, that's probably more then, or at the very least close to, the median price of a plain consumer laptop. Pretty sure it's even more true for the median price of a plain consumer desktop. Counting the drive that's already in the laptop or desktop.

  • Wow.
    That's amazing.
    And it would be six terabytes if you could squeeze the same density on a floppy.

  • If windows ever asks if you want it to check out the disk, just say no or be prepared to walk away for a week.

  • I've had 4 terabytes of storage now for quite a few years and my actual total usage seems to have peaked at around 2 terabytes about a year ago. It hasn't changed much since.
    • Interesting, but probably not a valid statistical sample size, unless you're very, very large. Also, I may have my units mixed up.

    • I would rather see access speeds improved. When they can deliver a 2 or 4 TB solid state drive at a reasonable price, then they can work on increasing sizes.

  • Yeah f'ing right (Score:5, Insightful)

    by rgbrenner ( 317308 ) on Tuesday May 22, 2012 @06:12PM (#40081243)

    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.

  • by Crypto Gnome ( 651401 ) on Tuesday May 22, 2012 @07:10PM (#40081775) Homepage Journal
    The "60TB" is actually an "up to" number.

    HAMR has a theoretical areal density limit ranging from 5 to 10 terabits per square inch, enough to enable 30TB to 60TB 3.5-inch drives and 10TB to 20TB for 2.5-inch drives

    From previous article about this tech from Seagate [].

    In reality do not be surprised to see 10TB and maybe 20TB 3.5 inch desktop drives in this timframe, but I for one WOULD BE surprised to see 40TB let alone the "in theory" 60TB.

    Having said that, I'd be extremely happy with a 10TB desktop drive.

  • by Osgeld ( 1900440 ) on Tuesday May 22, 2012 @08:53PM (#40082397)

    SSD's can be nice and fast but shit they are still pricy, and they have their issues ... coming from someone who still uses a old 80 meg scsi drive frequently on his vintage computers, I really dont want something thats going to kill itself in less than a decade

    mechanical drives are peaked right now in terms of speed, and on my main computer, with a ton of games on it, 3 OS and more personal files on it than I would ever use (projects and whatnot) am only using about 150GB (out of 500)

    great you can eventually slam 60T on a drive, maybe by then my 4T NAS would be full from me and the family, that is if all of our computers didnt have a half T in them already (8T if I combined it all into one resource)

    I dont store every single thing I have ever consumed for life on these things, and its going to be much longer than 2016 before I have a NEED for them, though at some point its futile to find a small drive for a reasonable price so they got us on demand ... I just want faster mechanical disks, something that can actually peak out a simple SATA1 Interface

    we get bigger, we get awesome interfaces but nothing to put on them other than overpriced, large ... for like 7 years ago, flash memory that slowly eats its own brain

    surely we can do better than just increasing space

Real Programmers don't write in PL/I. PL/I is for programmers who can't decide whether to write in COBOL or FORTRAN.