Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×
News

Ferroelectric Storage Density Tops 20KDVDs/Cubit^2 305

Posted by chrisd from the laser-beams-are-cool-too dept.
DeAshcroft writes "As reported in Technology Research News, researchers from Tohoku University, the Japanese National Institute for Materials Science, and Pioneer Corporation have demonstrated a prototype ferroelectric (as opposed to ferromagnetic) storage mechanism with density of 1.5 trillion dots per square inch. No word on why Japanese researchers are using square inches, but the new storage benchmark is the DVD. This is 47 DVD's in a square inch, or over 20KiloDVD's per square cubit. Original paper appeared in the Applied Physics Letters."
In related memory news, an Anonymous Coward writes "It appears the the ever present pause between photo's on a digital camera might finally be fixed. A company now claims http://www.mobilemag.com/content/100/102/C1396/ ) to have kicked up the write speed on a compact flash card up to 4MB/sec. This means we lesser photographers can now get the right action shot just by volume alone ;-)"
This discussion has been archived. No new comments can be posted.

Ferroelectric Storage Density Tops 20KDVDs/Cubit^2 More

Ferroelectric Storage Density Tops 20KDVDs/Cubit^2

Comments Filter:

  • Those Japs love their cartoon porn. (Score:3, Funny)

    by Anonymous Coward on Thursday January 16, 2003 @02:32AM (#5092770)
    Yeah, who else who measure in inches except the US and porn stars ?
  • At least right now what type of applications would this be good for? Do we really need that much storage? Perhaps if programmers wrote better code........... Then again remember when 2megs of memory was "the bomb" ?

    • Re:Is this really important? (Score:5, Insightful)

      by isorox ( 205688 ) on Thursday January 16, 2003 @02:37AM (#5092786) Homepage Journal
      Do we really need that much storage?

      $GENERIC_QUOTE_640K

      Video takes up a lot. Try storing multi-channel (multiple camera angles) uncompessed HDTV, gigs soon add up. Mix in some form of holographic projection and a dash of libraries of congress and you eat up terrabytes.

    • Re:Is this really important? (Score:4, Insightful)

      by Max Romantschuk ( 132276 ) <max@romantschuk.fi> on Thursday January 16, 2003 @04:42AM (#5093109) Homepage
      Most people don't. Anyone doing video or audio stuff do.

      I make music in my spare time. All the source files, sample libraries and raw audio of one single track won't fit on a CD anymore, even though the final track will only be something like a five megabyte mp3/ogg.

      Storage is like money, if you have enough you don't think about it too much. That way you end up with more time to do what you really want to :)
    • full 4d graphics for a 3d representational system. I'm not talking about simple computational 3d models or 2 layer 3d monitors, I'm talking about fully 3d displays and the data needed to display a full model on them...no matter how many voxels you use, you need a shitload of data.
    • > At least right now what type of applications would this be good for? Do we really need that
      > much storage?

      Yes, we really need that much storage. I am currently working on a data mining/analysis project. We currently are looking at multi-terabyte datasets from astro and climate simulations.

      Three dimensional super nova explosion simulations take around 50 terabytes at the current resolution. This is small compared to the petabytes of data needed for biological simulations.

      Look at http://sdm.lbl.gov/sdmcenter/tasks.htm [lbl.gov]for some discussion of why we need huge storage capacities (and some of the other problems involved).
    • At least right now what type of applications would this be good for? Do we really need that much storage? Perhaps if programmers wrote better code........... Then again remember when 2megs of memory was "the bomb" ?

      Yes, it is important. Because applications won't grow to fit the need if there is no room to grow. Yeah, I do remember when 2MB of memory was the bomb, I paid an extra $200 for that extra MB when I bought my 386DX-33. I had the best hard drive too, 80MB. Now I have 3x+ that in memory.

      I see your point, but look where things have gone in the last ... damn, has it been 12 years already?! I don't think you can blame programmers for writing worse code. Look at what the code of today is capable of, versus the code of 1991. Wolfenstein 3d vs Return to Castle Wolfenstein.

      No, we don't need the space right now, but we will find new and interesting ways to fill it if it is there. Imagine not having to uninstall your OS, just create a new 100GB partition, install the new OS to it, and boot to that one instead of the old one. We have gotten used to having to uninstall software because we have limited space to deal with. Think of all the things we wouldn't have to do if we had "unlimited" nearly instant-access disk space.

    • At least right now what type of applications would this be good for?

      KaZaA (Lite)

      Do we really need that much storage?

      Average movie: 800MB

      Perhaps if programmers wrote better code...........

      Perhaps if pigeon hole theory didn't exist, we could compress everything down to one byte and it wouldn't matter.

      Then again remember when 2megs of memory was "the bomb" ?

      No, I remember still having to be conservative with memory usage and what I could do with it. Preloading? Don't think so... now playing memory tricks such as precaching in code is possible. Before hand, it wasn't. 2MB was maybe the bomb for people who didn't code anything past a text editor. Even then.. don't edit large documents.
    • Some other responses have mentioned video, but let me add some real numbers to that. I'm a technician doing customer service repair on storage systems for high end digital video production systems.

      One of our high end systems uses 10 73GB drives. Yes I know that there are bigger drives out there, but for video bandwidth, and therefore rotation speed, is very important, as is stability. Tons of storage space is pretty meaningless if there are hiccups in your video stream.

      Anyway, these are arranged in 2 RAID-3 LUNs, which basically means 8 drives for storage and 2 drives for parity (error-correction). That gives 584GB of storage, which translates into just under 25 hours of high quality standard definition (NTSC) compressed (MPEG-2) video with stereo sound. How's that for eating up storage?

      At the moment I don't deal with HDTV, but I would estimate of the cuff that HD would cut that space down to the 8-10 hour range. Even though my company now offers a solution based on 181G drives (for about 1.5TB of storage!) that only brings that up to around 24 hours worth, which really isn't much when you consider that for many productions one hour of source material for 2 minutes of final product is considered a pretty good ratio. That means that 1.5TB could be eaten up with just the source material for a single one hour show!

  • What?!? (Score:3, Insightful)

    by trotski ( 592530 ) on Thursday January 16, 2003 @02:36AM (#5092781)
    They measure storage density in DVDs now?!?

  • shz/in^2 (Score:5, Funny)

    by violently_ill ( 629903 ) on Thursday January 16, 2003 @02:38AM (#5092789)
    sure, that sounds like a lot of storage, but how many full-length german sheizer films can it hold? when will we start setting standards that are actually meaninful?

  • Cubit^2 (Score:5, Funny)

    by GMontag ( 42283 ) <gmontag@guymontag. c o m> on Thursday January 16, 2003 @02:39AM (#5092793) Homepage Journal
    Cubit^2

    This sounds like an achievement of biblical proportions!

  • Cubits? (Score:5, Funny)

    by MrYotsuya ( 27522 ) on Thursday January 16, 2003 @02:40AM (#5092795)
    What the hell, is God telling them to build an ark?

  • Inches? Cubits? (Score:5, Informative)

    by cperciva ( 102828 ) on Thursday January 16, 2003 @02:41AM (#5092799) Homepage
    For those of you using sane units, this is about 250 gigabits per cm^2.

    • Re:Inches? Cubits? (Score:5, Funny)

      by Twirlip of the Mists ( 615030 ) <twirlipofthemists@yahoo.com> on Thursday January 16, 2003 @03:23AM (#5092910)
      For those of you using sane units, this is about 250 gigabits per cm^2.

      That's 2,412.1 petanybbles per acre, for those of you who prefer units with a little character. ;-)
      • What is a nybble? I know about nibble (4 bits, right?) and nipples (2 bits ;)). Is it a simply misspelling or is it a real unit? It is kind of hard to know in these times when we get idiotic crap like MibiBytes (Yes, I have 671.08864 MibiBytes in this computer... need I say more?).

        And to second a bunch of other people, stop with the inches. I seriously have a hard time taking someone scientificly serious when they don't use SI units today. And why are we still using inches for speakers and tires still?

        But regardless of the dumb units, being able to save all that on a single unit (disc?) means I could get uncompressed video in high resolution with uncompressed audio. But I guess they will put something even dumber than regions on it and screw it up.

      • Re:Inches? Cubits? (Score:4, Funny)

        by robbyjo ( 315601 ) on Thursday January 16, 2003 @03:55AM (#5093017) Homepage

        If you want a little character, that should be 2.035e-2 Library of Congresses per cm^2. ;-)

      • Re:Inches? Cubits? (Score:4, Funny)

        by Alsee ( 515537 ) on Thursday January 16, 2003 @06:55AM (#5093382) Homepage
        Or about 70,000,000,000,000,000,000,000,000,000,000 bits per square parsec.
        That's 31 zeros.

        -
      • The full specifications are:

        Storage: 195.7 decabytes per 3.3 twips

        Rotational velocity: 948 furlongs per fortnight
        Weight: .5 a small child per coffee-table

        Operating temperature: As warm as a large cow.

        Operating noise range: Approx that of a dyslexic peg-legged mime performing an impromptu street rendition of Macbeth using a small matchbox containing 12 red ants as his only prop.
    • Originally, a cubit was meant to be the length of your forearm (from the elbow, to the base of the hand).

  • Read speed a bit low (Score:3, Interesting)

    by renoX ( 11677 ) on Thursday January 16, 2003 @02:42AM (#5092803)
    I hope that they could use several head in parallel at the same time to increase the reading speed and also (why not?) the writing speed.

    If I remember well, a company has already done this for CD-ROM, it was reading several track at the same time, they had a commercial product but I don't know if it sold well.

    I wonder why it hasn't be done with HDD?

    Note that I'm not talking about multiple heads (too expensive), but using one head to read/write several tracks at the same time.

    • That would never, ever work with magnetic hard drives.
    • It was Kenwood that made this CD-ROM, and it didnt sell at all. I believe Maximum-PC gave it a shitty rating too, the drives kept breaking.

    • Re:Read speed a bit low (Score:2, Interesting)

      by Eccles ( 932 )
      If I remember well, a company has already done this for CD-ROM, it was reading several track at the same time, they had a commercial product but I don't know if it sold well.

      That would be Kenwood Tech [kenwoodtech.com] and their TrueX drives. Seems a nice idea since they probably don't sound like an airplane taking off like other high-speed CDs do, but they had a high failure rate on their first ones that I don't think they ever lived down. It's too bad they didn't license the tech rather than trying to build it themselves.

      You can find their TrueX pages on google, but their home page announces that they've stopped making the drives.

  • The day will come when some people will record 360 degree video and every sound that happens around them all the time. No need for discussion about what happened, just replay it.

  • My DVD... (Score:5, Funny)

    by istartedi ( 132515 ) on Thursday January 16, 2003 @02:48AM (#5092818) Journal

    ...gets 40 rods to the hogshead, and that's the way I like it.

    Really now, the Japanese are using square inches because Americans know what a square inch is, and they do a lot of business with the USA. Seems pretty obvious to me.

    Also, they just happened to reach a "milestone" of 1.5 when measured in square inches. 1 square inch = 6.4516 square centimeters, so this is only about 0.235 per square centimeter. Maybe they should have a press release at 0.3/cm^2. But if it's less than 1, it's just not very good.

    To resolve this issue, I propose the introduction of a new unit based on the meter and corrected by a factor based on Moore's law or whatever it is that governs storage density. The correction factor should be adjusted to allow for press releases oh... say... every 3 months so that stock traders will have something to speculate about. I propose that the new units be called "Horcs" in honor of no particular person, place or thing.

  • Can it be converted into furlongs per fortnight? Well.. maybe LOC's per Furlong^3. And then you can estimate how many fortnights that will keep you up watching pr0n.

  • Libraries of Congress (Score:5, Funny)

    by RollingThunder ( 88952 ) on Thursday January 16, 2003 @02:56AM (#5092840)
    Anyone happen to know what this is in standard measurements - as in, Libraries of Congress per 8.5"x11"?

  • Simpler units (Score:5, Informative)

    by XNormal ( 8617 ) on Thursday January 16, 2003 @02:57AM (#5092842) Homepage
    The recoring area of a DVD is 14 square inches. So the density of this new recording technique is 14*47=658 times greater than a DVD.

  • Transfer speed? (Score:5, Funny)

    by tigress ( 48157 ) <rot13.fcnzgenc03@8in.net> on Thursday January 16, 2003 @03:06AM (#5092866)
    What I'd be interrested in is knowing how fast it reads, preferably in another sane measurement, like 8" floppies per forthnight.

    • They got you covered. (Score:3, Informative)

      by autopr0n ( 534291 )
      Using the prototype, the researchers were able to read 25 kilobytes, or thousand bytes, of data per second, said Cho. This is relatively slow -- it would take 10 seconds to retrieve a 250-page book at that speed, assuming 1,000 characters per page. It is possible to increase the read speed to 3.75 megabytes per second, said Cho. This would make it possible to retrieve the information contained in about 150 books in 10 seconds. Current disk drives have read speeds of about 20 to 50 megabytes, or million bytes, per second.

      So about 36 novels/hour.

  • The really interesting bits, no pun intended (Score:5, Interesting)

    by Powerdog ( 106510 ) on Thursday January 16, 2003 @03:10AM (#5092876)
    First, they're only currently able to read 25kB/s. Yes, 25 kilobytes per second. They think they can bump up the read speed to 3.75MB/s. But it's the write speed that's curious. The prototype writes at 2.5MB/s, and they estimate they can bump it up to 125MB/s. A medium we can write to faster than we can read!

    Second, their goal is 667 terabits per cm^2. Yep, about 2667 times more dense than the 250 gigabits per cm^2 they're claiming.

  • Confused which cubits where used (Score:3, Funny)

    by bartjan ( 197895 ) <bartjan@vri e l i n k .net> on Thursday January 16, 2003 @03:13AM (#5092884) Homepage
    Are those cubits persianroyalcubits, northerncubits, irishcubits, greekcubits, hebrewcubits, homericcubits, olympiccubits, sumeriancubits, egyptianroyalcubits, blackcubits, shortgreekcubits, biblicalcubits, egyptianshortcubits, romancubits, assyriancubits or hashimicubits ??

  • many applications (Score:2, Interesting)

    by Anonymous Coward
    It's not so much the amount of memory that is important but about how small a package we can fit it in. This will allow tablet PCs and other PDA styled devices to have what the desktop PC world takes for granted. Also imagine the costs savings from application like outer space computing (every pound going up costs a fortune right?). Give people time and you'd be surprised what uses they'll come up with.

    Me, pr0n.
    • to make it able to be considered for secondary storage they have to increase the read speed.
    • This will allow tablet PCs and other PDA styled devices to have what the desktop PC world takes for granted.

      <sigh> I'll sure miss the whine, heat and gyroscopic resistance of my HDD, though. Oh, and the clatter of thrashing heads, and the rare treat of crashing heads. These are the days my friends. Remember them well.
    • Have to be worried about sensitivity to damage - your average CD is reasonably resilient to scratches (provided they're small and your EC is good) but once you start really pushing the data densities a spec of dust could occlude gigabits.
      May have to bring back caddies - new hermetically sealed vacuum caddies...
  • How fast is read/write access? I'm guessing it can't be less than 0x1B CEN TC224 sets per femtofortnight.

    -B

  • Ferroelectricity (Score:3, Interesting)

    by Bruce Perens ( 3872 ) <bruce@perens.com> on Thursday January 16, 2003 @03:47AM (#5092978) Homepage Journal
    I'm not familiar with ferroelectricity. It sounds as if the pickup device is like a scanning tunneling microscope, with an AC field on the point to measure the impedance of an adjacent ferroelectric domain. They claim that there is a change in the impedance of the domain depending on its electrical orientation, and that they can flip these domains electricaly (presumably with the same device that reads them, I guess by putting a higher voltage pulse through it). They claim that ferroelectric materials are piezoelectric, and that they are distorting the crystal lattice to store information.

    So, is all this for real?

    Thanks

    Bruce

    • Re:Ferroelectricity (Score:4, Informative)

      by Anonymous Coward on Thursday January 16, 2003 @08:12AM (#5093586)
      Ferroelectricity is quite analoguous to ferromagnetism (they borrow the name even, there is no iron in typical ferroelectrics). Ferroelectric is a subset of piezoelectrics; ferroelectric's polarization can be switch by either applying an electric field or mechanical stress. OTOH, piezoelectrics' polarization is unchangeable. Piezoelectric effect refers to the behavior when the material is under stress, charge appears in the surface of the material (piezo comes from pressure in Greek). It sounds like they are inducing 90 deg domain switching as storage. One problem I can already see with this is fatigue; because of polarization of the material, there is a dramatic change of sizes of domains undergo 90 deg switches (on the order of 0.001). So they either got a material that is very soft, like PVDF or something like that, or the macroscopic behavior of all these bits just even out the stresses.
  • This means we lesser photographers can now get the right action shot just by volume alone ;-)

    now there are two problems with this:

    1) without knowing how to set up exposure / apature / focus / whitebalance / and most importantly - general image composition - your chances of getting the right shot does not improve all that much with volume.

    2) moreover, assuming that you are an amature and want to actually have some control over your images afterwards (i.e. store it in a raw format so you can adjust the exposure and stuff - btw it's very useful I seriously encourage anyone who can spare the space to do this), the transfer-rate is worefully inadequate. one raw 3Mpix image is usually some 6-9Mbytes; that's some 1.5-2 seconds per image. Considering that in the real world, 3fps is considered mediocre, continue to rely on your SDRAM buffers (for the ones who have the luxury of them) for a little longer. For JPEGs, high quality 2Mpix is still usually about 1M a piece - so don't expect too much if your camera actually have a decent pixel count.

    anyway - not that it's not good news, or that I am humorless and don't get the joke - but I think microdrives will be around for a little longer for good reason.

  • Density expanded (Score:4, Informative)

    by dietlein ( 191439 ) <(dietlein) (at) (gmail.com)> on Thursday January 16, 2003 @03:55AM (#5093015)
    Ferroelectric density: 1.5Tb/in^2
    8 bits to a byte -> 187.5GB/in^2

    Hitachi's (formerly IBM's) 180GXP line packs 60GB to a platter. According to their data sheet [hgst.com], that is 45.5Gb/in^2. Convert to GB, and we have ~5.69GB/in^2.

    When common HD technology reaches Ferroelectric technology, we'll have about 6TB in a top-of-the-line IDE drive.
  • "One of the earliest types of measurement concerned that of length. These measurements were usually based on parts of the body. A well documented example (the first) is the Egyptian cubit which was derived from the length of the arm from the elbow to the outstretched finger tips. By 2500 BC this had been standardised in a royal master cubit made of black marble (about 52 cm). This cubit was divided into 28 digits (roughly a finger width) which could be further divided into fractional parts, the smallest of these being only just over a millimetre."

    And someone reports the information density using this kind of a measure?

    Don't get me wrong, I'm not trying to flame here, but without reading the article, due to the very same reasons, I'm willing to bet my right testicle that it's made by US "scientist".
  • In marketing, bigger is better. Any unit which produces higher numbers therefore is better.

    That's why we'll be using horsepowers and not kilowatts for car engines forever.

    This also explains the use of cubic inches in this text.

  • Cho Lab Homepage (Score:5, Informative)

    by mattr ( 78516 ) <<mattr> <at> <telebody.com>> on Thursday January 16, 2003 @04:44AM (#5093113) Homepage Journal
    His lab is here [tohoku.ac.jp]. Please try to stagger your access so you don't slashdot him.

    The Japanese side of the main Phonon Device Lab has pdf'd scans of newspaper articles from September 10. The Japanese also uses 1.4 Terabits/sq. inch.

    A drawing on the bottom of this page [tohoku.ac.jp] shows that his ultimate goal of 4 Petabits/square inch is based on a bit being stored in a 0.4 nanometer square, the size of one BaTiO3 crystal.

    Interesting experiment [tohoku.ac.jp] on his page tells you in English how to make piezoelectric ceramics(in collaboration with Washington U.).

    It looks like there are a whole raft [washington.edu] of people from Tohoku U. at U. Washington doing nano-bio research, mems, piezoelectrics.. maybe sq. inch came from Washington. Their Center for Nanotechnology [washington.edu] looks neat.
    I wonder if they were involved in this storage technology development.
    • His lab is here. Please try to stagger your access so you don't slashdot him.

      Ok, I'm planning on connecting at 16:29.27 GMT, please choose some other time. Thanks.
  • For those who have moved on from imperial measurement (and considering that DVDs are 3-dimensional objects) can someone give this measurement in DVDs/m^3 please?

  • I shudder to think how much good drool has slipped into my keyboard reading hundreds of similar headlines over the years.. Put it in a CD/DVD size (or smaller) format that is cheap, (re)writeable, and durable (Right now all our opticle formats leave something to be desired in this department). Until then STFU. I'm still waiting for a DVD[+|-]R format whose media costs less than $0.50 a disc and can be burned in less than ten minutes.

    So please, until something like this is available at Newegg spare us.

    Kthx.
  • We should use bizarre measurements in everything, just to mess with people. How about the speed of light expressed as 1798639627696 furlongs per fortnight?
  • but the new storage benchmark is the DVD. This is 47 DVD's in a square inch, or over 20KiloDVD's per square cubit

    Well that may be the benchmark your using but I want all my vapourware data storage anoucements in Libaries Of Congress please.

  • It may be impressive in the sense that ferro-electric might have decent seek and transfer times, but it's certainly not very impressive in terms of raw storage capacity. 47 DVDs in a square inch. Think of how thin the actual storage layer on a DVD is (not the plastic cover), and it seems pretty easy to reach that number or higher with DVD storage technology recnofigured into a cube of layers.
  • Of course the 20KDVD media is protected so it only plays in Region 1
  • I'd say:
    1 Molehill (Imperial) = 1 Mountain (Metric)

    Cheers,
    Jonathan
  • can somebody please tell me, that is how many dozen floppies per square yard? that is the native unit, i use in my brain.

Slashdot Top Deals

Whoever dies with the most toys wins.

Close