Slashdot Mirror
New 100GB Optical Disk From Taiwan
Alt173 points to this article from Taiwan Economic News , excerpting: "The National Science Council (NSC) said Sunday that a local research team has successfully developed a new optical disc that can hold more than 100 gigabytes of information.
The research team was led by professor Tsai Ding-ping of National Taiwan University. The new disc can store 150 CDs of favorite songs or an equivalent of 20 DVDs, Tsai said.
By using "near-field" optical technology, the 100-gigabyte disc stores more than any other similar product in the world. The super-sized disc will be used at home to store large movie or music files, according to Tsai.
The near-field optical technology also allows the bits of information on a disc to be spaced closer together to increase the disc's storage capacity."
Because it's denser, think of how much more a scratch will suck on this thing... when are we going to need error correction on these?
------
"And may your days be long upon the earth."
I hope that the burning rates will be higher than the actual CD-ROM burning speeds, because at 8x, it takes about 24 hours to burn 100GB!
Violence is the last refuge of the incompetent - Salvor Hardin
The article said it has about the same capacity as a recent article on holographic storage: http://sci.newsfactor.com/perl/story/17174.html
But what about its speed?
Lots of talk here about how this could be the "ultimate bootlegging product." On the other hand, if the movie industry is smart, this could be the "ultimate bootleg killer."
The movie studios are very nervous about internet piracy, but there's a good reason why the vast majority see movies in theatres and rent or purchase DVDs instead of acquiring bootleg VCDs. The simple truth is that low bitrate videos suck. They have motion artifacts. They have substandard audio.
They don't meet our quality expectations. A DVD is vastly superior. So is a 35mm print in a theatre. That's why Spiderman and Clones made over a hundred million dollars each in their first weekends, in spite of the fact that vastly inferior bootlegs were available "for free" on the internet.
As the electronics industry begins to retool their equipment from CDR manufacture to DVD-R manufacture, the movie industry is going to run into the same problem as the music industry -- they are going to be selling a $15.00 product that can be trivially copied perfectly onto a $1.00 piece of media. Over the next decade or so, as internet bandwidth increases, we will begin to see file-sharing of actual DVD images.
How can the movie industry make file-sharing of DVD images undesirable?
The answer is by providing something much, much better. Current "digital movies", as projected in theatres, provide a vastly superior image to DVD, and require approximately 70-100 gigabytes of storage space. The movie industry should be preparing to transition away from DVD to a new "super DVD" format that offers at least HDTV resolution, and most importantly, a big, whomping data rate that is completely impractical for internet streaming, and completely impractical for copying to DVD without downgrading the video quality.
Such a technology, available for the home, would quickly relegate DVD-quality recordings into the "low end" of video, at the same time that the price barrier on DVD-recording equipment falls through the cellar.
The industry should also realize that copy protection is worthless. It will always be broken, and the longer it goes unbroken, the more severe the market effect once it is broken. The real solution to the piracy problem of inferior bootleg recordings is the age-old tactic of the salesman. Offer a vastly better product, and your customers will follow.
Am I the only one that finds the comparisons to CD burners a bit lacking in perspective? For one thing, its not like they are taking a current CD and stacking it vertically until it can hold 100GB. Instead, they're going to increase the density and add a few layers to it. More density = more bits per disc rotation. A 24x CD Burner would burn 3.6MB a second, but a DVD at the same rotation would be like 3.6MB x 8 because the density is higher.
What Im saying is that a single layer version of one of these disks would get filled just as quickly as itd take to fill a CD (huge leap in the amount of data, though...), it might take twice as long if it has 2 layers.. and so on. Who knows, it depends on how the burning technology works.
My point is that the data rate of a CD is going to be incredibly slow compared to the data rate of one of these disks as long as the data density is higher. So stop comparing it to CD Burners. It is sort of like saying that an airplane would be many times slower than a car because it's so much heavier.
"Derp de derp."
"makes sense in theory if you know geometry?????" maybe if you understand geometry but not if you understand compression. here is an explanation of how compression really works. the article starts off on the right foot by bringing up the pigeonhole principle and bitspace:
according to the pigeonhole principle, to represent an arbitrary string of n bits, you need n bits. think about it, there are 2^n possible configurations of n bits so you absolutely have to be able to express 2^n possible pieces of data. that should be fairly clear.
so traditional lossless compression really works by rearranging these representations of the 2^n bits such that you can use shorter representations for things that you expect to see more frequently (based on patterns or perhaps just plain statistical frequency). (at this point, maybe look up basic huffman coding as an example of statistical compression techniques)
getting back to the zeosync thing, though:
the article then continues on by asserting that you cannot get back higher dimensional data from lower dimensions, but that you can get lower dimensional data from higher dimensions. this is true. however, it's not for free.
in a higher dimension, you actually have an infinite number of mappings for the same lower dimensional piece of data. for instance, the two dimensional point (1, 2) can be equally well represented as (1, 2, 0) or (1, 2, 192). developing a one to one mapping of a higher dimensional space to a lower dimensional space completely defeats the purpose of a higher dimensional space being used, since at that point a 2 dimensional representation will be identical to the three dimensional representation in terms of useful information.
now they claim they have a relational differentiation encoding technique that can represent a point that is both a square and a cube. this is not a big deal. let's say a 2 dimensional square (2, 2) is mapped into 3 dimensions. for example, we can choose to map it as (2, 2, 0) or (2, 2, 2). note that (2, 2, 2) is a cube, just as they predicted! wow!
somehow they claim that this ability will result in some savings when compressing, but the real problem is that (2, 2, 2) takes more space to store than (2, 2) and it is also now ambiguous what it means. you'd need to tack on another piece of information like how many dimensions to interpret the result as.
so in summary, at best, they are breaking even with the straight 2 dimensional representation of the data and at worst they are requiring additional space by using higher dimensions.
(yes, i know no one will read this post probably, but still, these zeosync guys were trying to convince people that their techniques would work by means of throwing around buzzwords. that's inexcusable but fairly typical of vaporware. they're trying to take advantage of the fact that it's unintuitive to think in dimensions higher than 3, so people will be less able to shoot them down.)
I can say I actually made a few bucks on their fluctuating valuation. They're now delisted. (Note that delisted != bankrupt, just that they ceased to meet NASDAQ's minimum requirements to let them borrow money from the public via stock offerings.)
But seriously, the RIAA/MPAA has not been the major block to their success. As of this year, they even had a major studio (Universal, I believe) looking into the technology, potentially even for the next DVD standard.
Their problems have been entirely related to cash flow (bring up a price chart for CDDD, if you still can- investors fled during the collapse of the 'bubble,' and there's even a rumor about a really rich SOB losing his margin account, leading to the huge selloff last year), and their engineering has been further hindered by the situations in Israel, and the current state of US-Russia relations. Unfortunately, this means that despite the multibit read tech, their demonstrations haven't broken any speed barriers yet (IIRC, their last demo was reading at about 2.5MB/s; I could be quite off, so check the press releases).
They still have some VC money (and a renewed funding agreement), but it's definitely a case of a good company losing due to a bad market. Hopefully they'll pull through and have some hardware to show before all this new stuff totally eclipses their tech. I'm a fan of alternative storage (viz the magneto-optical drive in my machine), and the fluorescent tech promises to be a bit more scratch-resistant than competing ideas (a big concern to me personally, as I'm the type who leaves all his CDs stacked next to the stereo.)...
...and I'm just posting AC to keep my Slash-addiction in check. No conspiracy at work, honest.