New Technique For Optical Storage Claims 1 Petabyte On a Single DVD
melios writes "Using a two-light-beam method a company claims to have overcome Abbe's Law to dramatically increase the storage density for optical media, to the 9 nm scale. From the article: 'The technique is also cost-effective and portable, as only conventional optical and laser elements are used, and allows for the development of optical data storage with long life and low energy consumption, which could be an ideal platform for a Big Data centre.'"
I was wondering where my pron collection would fit...
Supersnore. It's another year and another story about 1000-sublayer thick DVDs using multispectral lasers to fit ALL the DVDs on it. But how many of those make it to market? How robust is it? How much does that media cost?
I've been reading stories like this for 20 years and I still get little-girl-meets-Bieber excited when I think about being able to back up to just one disk... But it never happens. Spinning rust remains the cheapest and most convenient mass-storage device.
I have been looking forward to this for a long, long time... I
As screw the build up. I just wanted to say "peta-file" It's a funny word.
it will take about 1million seconds to fill it or about 11.5 days
Now our politicians and bankers can leave even more customer information on a train.
Sure, it's pedantic, but a DVD holds only about 9GB on a dual layer disc. It may be a disc with the same form factor, but it's not a DVD. You could just as well say "holds one Petabyte on a single CD", which also wouldn't be true.
I was promised a flying car. Where is my flying car?
This is great news for ASICS. Maskless direct write is the holy grail for this. Most of the cost of IC making now lies in the mask set and cost 10's of millions of dollars for a top line chip. There are ways to 'double up' mask steps into one reticule to save money on medium volume ICs and small volume has to be done on MultiProjectWafers.
Direct write is slow but with a multiple beam setup that can be speeded up. I'm thinking what Mapper Technology is trying to do with e-beam.
I've just about had it with all optical media in general. I've had numerous CDs and DVDs over the years that just stopped being readable without even having any visible damage. Both self-written and factory discs. I'm only halfway through re-watching a retail set of Stargate SG-1 DVDs I purchased at Costco for $179.99 just 3-4 years ago, and I've already encountered a handful of discs with serious defects. Learned my lesson not to buy physical media anymore. Once I finish torrenting a good pirated version of the series I'll probably never try to watch the DVDs again. The box is nice though.
Bottom line is even if one of these amazingly high density optical media schemes finally pans out, the media will need to be composed of pure diamond or something else incredibly durable, and have a filesystem with incredible levels of error correction and redundancy or it will be pointless to put even a terabyte of data on such a disc, much less a petabyte. And that's not even bringing up read/write speeds and other issues that have already made this type of media useless for many purposes.
Those clowns at InPhase ("Holographic Discs") were like the Duke Nukem Forever of storage; well over a decade, and no shipping product.
For now, I put this in the same pile as the Windows Database File System and Laptop Fuel Cells.
The "one beam going through a doughnut beam" technique is well known through STED microscopy: http://en.wikipedia.org/wiki/STED_microscopy
STED is a superresolution technique for imaging when using fluorophores.
This is a very nice idea using the technique in a different way for a different application.
Do you have any better hostages?
With even home Internet providers enforcing monthly caps, how will you fit your 3D 4K movies across a home Internet connection without having to take a week off surfing after streaming a single movie?
With this new technique, Netflix can just actually send you 1 disc with *all* the movies on it...
It must have been something you assimilated. . . .
for video, for a given bit stream size, are you better off with higher resolution and higher compression, or lower resolution and lower compression?
That depends. Downscaling and compressing at lower resolution is in theory equivalent to running a blur filter over the whole high-resolution picture and then compressing the blurred high-resolution picture. But in practice, video codecs prefer to handle features of a given size. MPEG-1 and MPEG-2 use 8x8 pixel cosine transform blocks, as do H.263-style codecs such as Spark, ASP (DivX), and Theora. H.264 and VP8 use variable transform block size to efficiently handle both flat areas and detailed areas in the same picture. Video codecs also differ in what sizes of motion they compensate for. Some codecs support precision down to the quarter or even eighth pixel, while others support only half pixels. Other codecs aren't effective at 1080p HD because their maximum displacements aren't big enough.