Slashdot Mirror
Holographic Storage Crams in 0.5TB Per Square Inch
An anonymous reader writes "VNUNet is reporting that a company called InPhase Technologies claims they have successfully recorded 515GB of data per square inch to capture the record for highest data density. From the article: 'InPhase promised to begin shipping the first holographic drive and media later this year. The first generation drive has a capacity of 300GB on a single disk with a 20Mbps transfer rate. The first product will be followed by a family ranging from 800GB to 1.6TB capacity.'"
I was looking for some details on the storage mechanism and specifications of data decay, reliability and such, I didn't see anything on that. Will normal error correction be sufficient for such a device, or does it make sense to use the same disk to write every bit of data onto it more than once in different locations, say 3 times alltogether and when reading, compare the bits and chose the value that happens at least 2/3 times? Will data decay on this media any faster or any slower than on a normal magnetic disk?
You can't handle the truth.
For high output data rate, one must read holograms with many pixels per page in a reasonably short time. To read a megapixel hologram in about 1 ms with reasonable laser power and to have enough signal at the detector for low error rate, a diffraction efficiency around eta = 3 × 105 is required. To write such a hologram in 1 ms, to achieve input and output data rates of 1 Gb/s, the sensitivity for this example must be at least S'eta2 = 20 cm2/J.
Since this hologram was retrieved using a readout pulse of 1 ms, this experiment implements the optical signal (but not the subsequent fast electronic readout) of a system with a readout rate of 1 Gb/s.
They are using optical storage technology, not terribly dissimilar to CD-R and DVD-R technology.
So, how well do their disks stand up against bit-rot?
If it's holographic, aren't you more concerned with how much data per cubic inch? I mean, per square inch kind of loses meaning at that point, doesn't it?
And why is 1.6TB the largest they're offering? 3 square inches of recording surface? A 5.25" drive gives what, about 70-80 square inches of recording surface? Give me a 40TB drive for the price of a 500GB hard drive. That'll be worth something to me, otherwise I'll just stick with standard hard drives. They're cheap and fast.
More promises, no product move along.
Back in 2000 or 2001 slashdot had a story about a company called C3D (or CDDD which was their stock ticker, website was http://www.c-3d.net/). This company promised 1TB and higher density discs with insane transfer speeds because it was storage...in 3D. They showed a few discs (CD sized) and a reader which were supposedly a prototype of some sort at trade shows. All of this ran their stock up quite a bit. They were promised to replace DVD's in a few years, and eventually hard drives. There was also this credit card device (10gigs) which was rewritable (?), which was to replace traditional hard drives in notebooks.
Deadline after deadline passed, the stock slowly declined ($60 a share was the norm in 2000) due to the market conditions in 2001, eventually causing it to be delisted from the NASDAQ (has a value of $0.01 a share). Rumor has it that the company was founded/owned/something by a former Israeli/Soviet general (the company wasn't located in the US), and that there never was a product (all demos were faked).
How do I know this? I was the fool who bought the stock when it was $20 a share, watched it rise up to $66, and fall to nothing. I believed before and it cost me a decent amount of money.
Holographic media has been a scam before and it'll be one until there is a box with a price tag in a store. Even then, I would be cautious about buying it.
"Never underestimate the bandwidth of a station wagon full of tapes hurtling down the highway."
...or whatever the exact quote is, as I couldn't find a reliable source for it.
~ Andrew Tanenbaum
here I thought I was the only one who thought of that. The whole "per square inch" thing really only counts in terms of 2-dimensional media. Which brings up the question, How many bits/square inch does this give? That is, how much per square inch, single layer? And just how many layers does this employ?
The difference between spam and poop is that you don't have to dig through septic tanks looking for real food. -- Me
You *really* must have meant 1.6 *TB* instead of GB. Then your numbers would make sense.
1.6 GB * 8 * 1024 = 131107.2 Mb - megabits in 1.6 gigabytes
131107.2 Mb / 20 Mbps = 655.36 s - seconds to read at 20 megabit per second
655.36 s / 60 s = 10.92 min - convert to minutes
At 20Mbps, it would take you 4.855 days to read a terabyte, which is pitiful for local storage. (1.6TB would be 7.77 days, or the almost 8 days in the parent post.) Even at 20MBps, that is still 14.56 hours for 1 TB, which is far too slow.
This might work as a backup medium for archiving, as long as it was suposed to be 20 megabyte/s instead of megabit. Many tape systems are right around the 20MBps mark, however there are solutions out there that archive over 100MBps.
File management on a terabyte plus drive will be a breeze to boot (to coin a phrase).
I imagine the "find" option in Windows will have no problem nor will Spotlight. And those wonderful desktop search tools will just FLY indexing a terabyte. No sweat.
That or I'll lose 6 out of 8 hours either organizing or just searching for 2k in 2,000 gigs / 2,000,000 mbs / 2,000,000,000 kbs.
Can't wait. Or we can all wait for that wonderful file system that's yet to come.
- on that note a serious question -
WTF happened to "that guy" who was working on a desktop-calendar hybrid model for a UI? It resolved entire workdays into desktop snapshots that were presented like a scrapbook which one could flip through like a titanic personal organizer. This has been why I've been handling project data within chronological folders lately because - unless I've had serious head-trauma - I can recall WHEN I worked on something. Names? Either the one's I put in on a whim or some crypo-garbage that the app assigns? The latter really make searches fun. "That Guy" was a blurb on some TechTV show ages ago. All I can recall is he was in the Bay Area, was Jewish, and was loaded with PHDs. Which religious prefs aside makes him about as common as the water in the Bay itself as a google ref.
Still, I think this is where we have to go in future UIs - now I just need to find out where "that guy" - went.
The Holographic model of Data Storage isn't really that "un-tested". While we might not know how it works (the under-the-hood understanding of it) -- we knmow that it does -- the Human Brain stores memory in a holographic model (read the Holographic model of the brain as proposed by Dr. Karl Pribram) and infact a renowned physicist by the name of David Bohm suggested such a medium for the whole of the universe itself. But that aside -- the beauty of a hologram is in it's ability to retain all of the data it stores even though the physical medium itself might be disrupted/reduced somehow. IIRC, the concept goes like this -- you can cut a hologram into smaller pieces -- but each of these would retain the whole image. There is possibly a certain level of "differentiation" that needs to happen before the validity of the data gets compromised...
What would happen if we invested heavily into developing factories to manufacture these, stuffed them into cheap durable nintendo style units, preload every book indexed by the google book project and every song we could lay our hands on and distributed them to every man, woman and child on earth?
-1 Uncomfortable Truth
As the subject line says, I'm very very worried. I mean this from a "1984" standpoint.
We've already read stories about how our past activities on the Internet (news groups, blogs...etc) can catch up with our future in a very bad way. With storage getting cheap and more abundant, I fear that giant archives of public data will be collected daily and stored for hundreds of years...all ready to be pulled for review later. Any place, at any moment, digital video of you recorded in public can be data-mined using facial feature algorithms to track your history of where you went, when, and for how long.
While such technology will certainly be available in the UK, there is nothing against US law from preventing it happening here. Homeland Security, Patriot Act...bla bla bla. It's just a matter of time when terabytes are cents on the dollar.
Life is not for the lazy.
Well I sort of agree with you about solid state storage but there are a few flaws in your logic.
I will answer with a question. Why do people or organisations backup their data?
Most PC uses may backup their data to floppy (now obsolete) or to zip (getting obsolete) or CD or DVD but rarely to tape because it is not cheap and it is inconvenient.
Most organisation backup their data (from a few 10's of GB to 1000's of TB) mainly with tape devices (super DLT's can save approx 400 to 600 GB per tape) with approx 50MB/sec through put. This solution can be very expensive ($10k to $100M). Solid state storage can do this but the cost would get quite expensive. Even disk storage would be cheaper than solid state however any company that wants to stay in business should have a Disaster Recovery Program for their IT department and this means off-site storage of backup media.
Government organisation (ie. Tax Department) require data retention for seven years and some even longer. Now solid state and even disk backups become so expensive you need to have the budget of NASA for even a small company. If the company requires off-site backups then at the moment tapes are the only solution.
What is important here is the potential of the Holographic Versatile Disk (HVD). They are starting with 300GB @ 20MB/sec and assuming the writers/readers are say $2k each (guessing here but reasonable) you could get a multi-stacker silo (say 6 heads and 50 disks) for say $20k that would have a through-put of 120MB/s which would be fine for small to medium sized companies. An equivalent tape machine would cost close to $50 and up depending on your tape silo and not only would it would be much larger it would have a slower response for the robot stacker.
Assuming a HVD multi-stacker library of similar performance to it's tape/cartridge equivalent the overall cost now comes down to the media and I am quite sure HVD disks will be at least 10 to 20 times cheaper than the equivalent tape/cartridge. Coupled with that will be the potential longer life and small storage area of the HVD, not to mention "near-line recovery" capabilities.
What we are seeing with HVD is a change away from tape backup units and anyone who has worked in this area will welcome that.
On an interesting note. It looks like the VHS tape recorders are being phased out by HD and/or DVD recorders. I am sure the same will happen when HVD are pushed as the new backup strategy. Think thin DVD's (even in a protective case) compared to bulky VHS cassettes.
Note: You should not compare CD's, DVD, HD-DVD and BluRay with HVD since they are aimed at different areas of IT, however that is a subject for another day.
There ain't no such thing as proprietary standards only proprietary formats. Standards are by definition open.