Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Data Storage

Ultra-High-Density HDDs Made With Graphene Store Ten Times More Data (phys.org) 62

Posted by BeauHD from the data-density dept.
Graphene can be used for ultra-high density hard disk drives (HDD), with up to a tenfold jump compared to current technologies, researchers at the Cambridge Graphene Center have shown. Phys.Org reports: The study, published in Nature Communications, was carried out in collaboration with teams at the University of Exeter, India, Switzerland, Singapore, and the US. [...] HDDs contain two major components: platters and a head. Data are written on the platters using a magnetic head, which moves rapidly above them as they spin. The space between head and platter is continually decreasing to enable higher densities. Currently, carbon-based overcoats (COCs) -- layers used to protect platters from mechanical damages and corrosion -- occupy a significant part of this spacing. The data density of HDDs has quadrupled since 1990, and the COC thickness has reduced from 12.5nm to around 3nm, which corresponds to one terabyte per square inch. Now, graphene has enabled researchers to multiply this by ten.

The Cambridge researchers have replaced commercial COCs with one to four layers of graphene, and tested friction, wear, corrosion, thermal stability, and lubricant compatibility. Beyond its unbeatable thinness, graphene fulfills all the ideal properties of an HDD overcoat in terms of corrosion protection, low friction, wear resistance, hardness, lubricant compatibility, and surface smoothness. Graphene enables two-fold reduction in friction and provides better corrosion and wear than state-of-the-art solutions. In fact, one single graphene layer reduces corrosion by 2.5 times. Cambridge scientists transferred graphene onto hard disks made of iron-platinum as the magnetic recording layer, and tested Heat-Assisted Magnetic Recording (HAMR) -- a new technology that enables an increase in storage density by heating the recording layer to high temperatures. Current COCs do not perform at these high temperatures, but graphene does. Thus, graphene, coupled with HAMR, can outperform current HDDs, providing an unprecedented data density, higher than 10 terabytes per square inch.
This discussion has been archived. No new comments can be posted.

Ultra-High-Density HDDs Made With Graphene Store Ten Times More Data More

Ultra-High-Density HDDs Made With Graphene Store Ten Times More Data

Comments Filter:
  • Amazing if they can really do 10X.

    • Maybe throw in a little helium.

    • Re: (Score:3)

      by marcle ( 1575627 )

      It'll be amazing, but it probably won't be cheap, or available to consumers. Don't expect this to show up on Newegg any time soon.

      • I'm sure the dudes mining Chia will be more than happy to pay premiums for these drives, especially if they are relatively fast and can handle a lot of wear cycles.

        • The wear and tear and ability to hold data for prolonged periods will make or break the technology.

          Theres still plenty of room for old fashion mechanical for the same reason tape hung around much longer than anyone anticipated;- Its good to have cheap and plentiful backup.

          a box of 10 or so 100 terrabyte drives backs up a small datacenter. A rack of those boxes backs up a big-ass amazon style one.

          • Exactly. There is always room for high capacity secondary storage, be it an external drive, a NAS, or a hybrid SAN or NAS. For example, backups always need to go to some type of random access disk array as a landing zone, just to reduce the time machines take to back up, as well as provide a copy of data for fast restore.

            The only downside about hard drives is the I/O bottleneck. This can be mitigated by having a smart array controller and a ton of SSDs, either as cache, or a direct landing zone for write

          • In fact, this kind of density may actually finally do away with LTO, if the drives are also resilient to bit rot.

            Unless I've missed something, LTO is still king of archival storage.

      • It'll be amazing, but it probably won't be cheap, or available to consumers.

        Like pretty much every brand new technology invented ever.

        • Yup but for the past decade or two they've stopped filtering down.

          • Re: (Score:2)

            by ceoyoyo ( 59147 )

            The $60 SSD in my computer says no.

            • Oh yeah so what is the capacity? Bet it is the same as the $60 SSD you could buy 10 years ago or barely increased whereas the actual storage density has increased by orders of magnitude.

              And of course, not much point in that SSD or high density storage without network links that can keep up with your multiple disk NAS right? How fast are those links? Oh right.. the same 1GB speed that was deprecated 15yrs ago? It is pretty sad when your crappy slow wireless outpaces your wired connection... even if the wirel

              • Re: (Score:2)

                by ceoyoyo ( 59147 )

                Bet it is the same as the $60 SSD you could buy 10 years ago or barely increased

                Seriously?

                https://images.techhive.com/im... [techhive.com]

                And the local tech shop across the street has lots of 10 GB switches that I can hook up to the fibre line that comes right to my wall for half the price, twenty times the bandwidth (and no cap) of the cable connection I had a few years ago.

                Maybe you need just need to upgrade your hardware?

                • Actual orders I placed for the industry standard and popular gear.

                  "SanDisk SSD PLUS 1TB Internal SSD - SATA III 6 Gb/s, 2.5"/7mm"

                  ORDER PLACED
                  November 26, 2018
                  TOTAL
                  $129.89

                  ORDER PLACED
                  March 9, 2020
                  TOTAL
                  $119.06

                  Today $97

                  Given Moore's law just the density increases alone should have halved the price every two years. These 1TB SSD's should $30-40 at this point.

                  And yeah there are 10GB switches you can buy... there were 10GB switches you could buy 15-20 years ago as well. They aren't consumer gear and the lower end

                  • Re: (Score:2)

                    by ceoyoyo ( 59147 )

                    but for the past decade or two they've stopped filtering down.

                    Bet it is the same as the $60 SSD you could buy 10 years ago or barely increased

                    Actual orders I placed....
                    ORDER PLACED
                    November 26, 2018

                    Lol. It's all right man. I expect you're getting to that age where the years kind of blend together and 2011 and 2018 are pretty much the same thing.

      • Yeah, hard drives haven't gotten any bigger or cheaper since 2002, when was buying the 80GB Maxtor EIDE drives for $259.

        Ten years ago, $150 would get you 750GB - 1TB. Today:

        https://www.newegg.com/seagate... [newegg.com]

        80GB or 8,000 GB - it's all the same.

      • Comment removed based on user account deletion

        • Re: (Score:3)

          by suutar ( 1860506 )

          At a guess, the parts of the drive that aren't platters cost a good chunk of $100 and so a price point lower than that would have essentially unsellable capacities.

  • Search Speeds (Score:4, Interesting)

    by theshowmecanuck ( 703852 ) on Tuesday June 08, 2021 @08:48PM (#61467856) Journal
    Just a random thought, given the vast amount of data these might store, how will they or how do they optimize the speed to get to any piece of data. Aren't the indexes getting proportionally larger as well? It seems to me that increasing storage capacity is just one side of the equation. I don't know disk design, just curious what they do to speed up read, never mind write for these things. I guess one other thing, if they could make these virtually frictionless, could they just put a read/write head in the form of a band from centre to edge directly in contact with the disk? No need to physically move a read head. I guess if there was vibration there wouldn't be any leeway to prevent seizing. I know these are naive questions but enquiring minds want to know.
    • I expect it is going to have to have a pretty large SSD attached to smooth out buffering, reads and writes.

    • Because hard drives drives are getting larger, but the I/O pipes are not expanding as rapidly (unlike NVMe SSDs which often are faster than the bus on a system), we may wind up seeing more mechanical tricks to get data out faster. One drive made about a decade ago had two completely different set of read/write heads, in an active/active configuration, so they could do completely independent transactions. Perhaps we might see that return.

      Form factors are also an issue. If the height of drives could be inc

      • Oh the irony. 5 1/4 drives that were popular most recently were technically 1/2 height. What's old is new again.

    • Just a random thought, given the vast amount of data these might store, how will they or how do they optimize the speed to get to any piece of data. Aren't the indexes getting proportionally larger as well? It seems to me that increasing storage capacity is just one side of the equation

      The space taken by indexes is not really considerably related to the size of the drive but more to the size of the file system IMHO. So, a 100 TB file system split across 10 drives will have the same file system indexes as a 100 TB file system on 1 drive.

      Similarly, if you put several smaller filesystem on one of those big drives, file system indexes should not be a problem. Nowadays, with virtualization, a host typically has several smaller file systems, at least one for each VM.

      There might be internal inde

    • Re: (Score:3)

      by AmiMoJo ( 196126 )

      HDDs don't randomly spread the data around like SSDs do, they store most of it in a predictable linear fashion. Then they have an a reallocation table that tracks the relatively small number of blocks that aren't where you would expect them to be.

      They use DRAM cache for the reallocation data. It also helps buffer data being written. SSDs use even more of it.

      • That is when it is first filling the disk. But after deletions and/or new files or additions to a file, data will go where there is space. Fragmentation.

        • Re: (Score:2)

          by AmiMoJo ( 196126 )

          Sure, but that's the filesystem. The disk has no idea about the filesystem.

          • But the heads have to traverse the disk to get to the files on the file system. If there are more places to search, the search takes longer.

    • Re: (Score:2)

      by dargaud ( 518470 )
      I seem to recall that during the late '90s there was a prototype CD reader with 100 heads at various distances from the axis. It would read it in a few seconds and store the content in a RAM buffer. I guess the internet put a stop to that, but the idea has probably been attempted. Now that I think of it, doesn't a fast HD that came out last week have 2 heads ?
      • I was just thinking that someone could maybe come up with a way where the head doesn't physically move, but electronically moves in some sort of solid state manner. That is, move where the magnetic field is up and down a solid bar. Then there is no actual physical movement. I Am Not A Physicist.

    • Spindle speeds have not increased at all in the 25 years or so that I have been dealing with hard drives.
      If anything, they've got slower on average as where you would previously have used a 15,000 RPM drive, you would now use an SSD.
      Storage density has got a lot higher, so if you have 100x more data per platter, running at the same spindle speed, you would get 10x faster data transfer - 10x more tracks, 10x as much data per track. A track takes the same fraction of a second to read regardless of the amount

      • Re: (Score:2)

        by grub ( 11606 )
        My first hard disk was 3600 RPM. Old 10 MB thing (forget the manufacturer) in 1984 or so. I remember the sales guy pumping up the fact that it could read "a meg a minute!"
        • I remember getting some disk compression tool at work and we were all excited that we could now store 10MB on a drive. lol.

          You had ones and zeroes? We had to make do with lower case L's. -- Dilbert.
          .

    • The good news about increases in aerial density of rotating storage is that given the same RPM, the higher-density drive has more bits moving under the head per unit time. Depending on the speed at which the head can actually read / set bits in the magnetic recording layer, increases in density also result in improved performance.

      Of course, this all goes to hell with a fragmented drive because you have to wait for various parts of the file to appear under the head(s) - this is one place where SSD always wi

      • I just assume disks become fragmented and that is my head immediately jumped to the problem. HDDs always become fragmented, and then the search is on. It's the nature of the beast.
    • There are a couple major problems that go along with ever increasing HDD (and SSD) capacities.

      1) The speed increases are much smaller relative to capacity increases. Every time the capacity of a device doubles, the read/write speed only seems to increase something like 1.2x. This means it takes longer and longer to fill up the drive or read all the data. If a drive in a RAID configuration fails, it can take days to rebuild. If a virus or ransomware attacks all your data, it can likewise take days to rest

  • Huh? (Score:5, Interesting)

    by Waffle Iron ( 339739 ) on Tuesday June 08, 2021 @08:57PM (#61467884)

    The data density of HDDs has quadrupled since 1990,

    OK, I just pulled up an archived image of the receipt of the first disk drive I ever bought: In 1991, I paid $399 (worth almost $800 now) for a 100 MB hard drive. Right now, Amazon is listing 4 TB drives (still with the exact same physical dimensions) for less than $100.

    That's 40,000X more dense. TFS is off by about four orders of magnitude.

    • Why are you using prices to compare aerial data density?

      Did you consider that perhaps that 4 TB drive has many more platters than the 100MB one did?

      • Re:Huh? (Score:4, Funny)

        by Waffle Iron ( 339739 ) on Tuesday June 08, 2021 @10:03PM (#61468022)

        I'm not using the prices; I just thought it was interesting.

        But you could be right; there very well might be 10,000 platters in the 4TB drive, each stacked only one micron from the next. Astounding!

      • Re:Huh? (Score:5, Informative)

        by Entrope ( 68843 ) on Wednesday June 09, 2021 @02:48AM (#61468376) Homepage

        Capacity growth of 40,000 times isn't going to be explained by 4x areal density increase and 10,000x platter count increase. Checking a few reviews, it looks like 18 TB hard drives typically have nine platters, compared to three platters [tomshardware.com] (and five heads) for a commodity HDD in 1991. So platter and head counts account for less than 4x growth since 1991.

        • Adding more platters means making thinner platters that are still well balanced for high speed rotation, and adding more heads as well, and a more robust motor to get that additional mass rotating, and that means adding cost to the bill of materials - more than just the cost difference of aerial density increases.

      • Re: (Score:2)

        by Bengie ( 1121981 )
        600,000,000 increase in density over the past 60 years, yet only 4x in the last 30? right. Not to mention the 40,000x reduction in cost over the past 30 years. I think they got the significant correct and missed the magnitude.

    • I'm struggling to understand that math as well. Could they have meant 4x in each direction, for a 16x gain? And maybe they didn't account for depth-wise gains at all, namely the massive improvement seen with the shift from writing parallel to the platter to writing perpendicular, into the platter? And the second part of that sentence is just as confusing:

      and the COC thickness has reduced from 12.5nm to around 3nm, which corresponds to one terabyte per square inch.

      From what the summary says, the COC is basically just the coating on the platter, which means it doesn't store any data at all. A reduction in the COC's th

      • The closer you can get the read/write head to the data layer of the platter, the more data you can store. The thickness of the COC is what determines that.

      • The further the head is from the platter, the more the magnetic field spreads out, so the lower the data density. And of course, the reverse is true. So get the head closer for higher data density.

    • Re: (Score:2)

      by ookaze ( 227977 )

      I paid nearly as much at that time for a 20 or 25 MB disk, which I believe still works. The difference is probably that mine was a SCSI disk for my Amiga computer.
      Like others have noted, it gives no indication of the data density on platters.

  • From the abstract, HAMR raises the disk to 800K for 1ns so there are worries about thermal stability against laser irradiation.. But they say 5 years drive life should be fine.
    I've seen drives fail in 2-3 months so 5 is good? Doesn't sound great but since it is at an early stage this is amazing.

    • From the abstract, HAMR raises the disk to 800K for 1ns so there are worries about thermal stability against laser irradiation.. But they say 5 years drive life should be fine.
      I've seen drives fail in 2-3 months so 5 is good? Doesn't sound great but since it is at an early stage this is amazing.

      I'm on my second WD NAS. The first was about 1Tb. The one I'm using now is about 3Tb and probably 6 or so years old. I basically use it for streaming to either my Kodi box or to my Roku TV.

      100Tb would probably do me for as long as the drive stayed alive. Backups would kinda be a pain though.

      I guess I'm more interested in longevity than speed for the most part. I would prefer to not replace my drives every 5 years, much less 2-3 months.

      I'm running a 1Gb network wired, and it works pretty good, even for 4k vi

  • Now all we need.. (Score:4, Insightful)

    by Z80a ( 971949 ) on Tuesday June 08, 2021 @09:18PM (#61467926)

    Is the mass produced graphene

  • But can they actually produce this at scale or will it be like every other amazing sounding graphene story, lab conditions / bespoke construction only.

  • Excellent (Score:1)

    by Anonymous Coward

    Another graphene breakthrough, I can use this to store data on my fusion powered flying car.

  • Graphene is quite amazing stuff with scores (hundreds?) of potential applications.

    Does anyone know of any actual product shipping or use in service yet for graphene?

    I'm not saying I think it is pie in the sky, but it is taking a long for it see real application it seems to me. I just looked up a list of commercial graphene applications, found a list of 60, but it was not clear if any of them were more than potential applications.

    • There's tons of theoretical applications, there is just no viable way to produce it outside a lab in a manufacturing environment and leverage it in other parts of the supply chain to make products and goods to date.

    • Oddly, there are jackets that use single-layer graphene to perform heat distribution. They're expensive and hard to lay hands on though, because I don't think there's really a good mass production process.

      Here's an example. [vollebak.com]

    • Re: (Score:2)

      by WoTG ( 610710 )

      I think these are in production, or maybe just available for OEM orders:

      https://www.martinrea.com/news... [martinrea.com]

      Graphene in brake lines, of all things.

  • Given a process not dissimilar to regatory capture is trying to make this site about politics, it's good to see a nice, throbbing nerd subject make it to the front page once in a while.

  • Let's see, we could take the platter from a hard drive, and make it interchangeable....

Slashdot Top Deals

1 Mole = 007 Secret Agents

Close