Slashdot Mirror
InPhase Technologies Promises Holographic Drive in May
Anonymous Coward writes "After 8 years of effort, InPhase Technologies is shipping the world's first holographic disk drive next month. They showed it at this week's NAB. With a 300GB 5.25" disk cartridge and a 50-year media life, the Tapestry 300r is aimed at the video and film archive market. They've been promising this thing for so long I'd given up hope that they'd ever ship it!"
I've dreamed often of the day I could buy a completely non-standard technology that rids me of large quantities of the pesky money I have lying around while at the same time solves the removable storage problems of 3 years ago. Too bad this unit only costs $18,000 and stores just under 1/3 of my hard disk space!
I'm a big tall mofo.
When I can buy it.
But I won't actually buy it until after I hear at least 1 horror story about photonic lifeforms eating somebody's data or something equally bad:)
$18,000 could buy me enough hd's so that i could rotate 2 backup disks once a year for the next 90 years.
If you mod me down, I will become more powerful than you can imagine....
Wow, that brings back memories of 5.25" floppies. Makes me wonder what this tech will look like in 20 years. Internal drives with 3.5" media storing hundreds of terabytes? SD-sized holographic media? Now that this technology has moved from proof-of-concept to a purchasable product (or will be in one month), it'll be very interesting to see how quickly it progresses.
Unfortunately, my hand passed right through it.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
FTA:
Spinning Hard drives, Solid State Hard drives, CD's, and DVD's don't have anything CLOSE to holographic media.
Spinning Hard drives could be used, and they are, to store data for long periods of time. Problem is that it susceptible to EM fields and even while not spinning, it might be possible to have some degradation nonetheless. Holographic media is not affected by EM fields.
Solid State Hard drives are better off than spinning ones for sure, but still suffer from the same problems with an EM field AFAIK.
CD's and DVD's long shelf life is a MYTH. Most of them are not manufactured to last longer then 5-10 years. A scratch can easily damage either one of them, and repairs are not easy. Holographic Medium? Apparently not.
So the
The fact they actually got it to production and selling it means there is a pretty good chance of seeing a few thousand dollar reader/writer within 2 years.
For those that are really hung up on the price, consider this:
To be REALLY safe with your data you would have remove all single points of failure. A single hard drive on a shelf IS a single point of failure, as is a CD/DVD. So you would need to be constantly "rolling" over the data in multiple RAIDS with snapshots, while at the same time, verifying the integrity with checksums before every snapshot. To take it one step further, multiple locations that synchronize over high speed networks... iSCSI?
Apparently a holographic medium can be written with "hundreds of holograms being stored in the same physical area". Sure sounds to me like you could store quite a bit of data with a considerable amount of recovery capability. I would hazard a guess, that just a few of these written this way and stored in separate physical locations would provide the same level of reliability and redundancy that current solutions provide (such as the one I outlined)... with a 50-100 year shelf life. If you look up the actual costs of iSCSI this sounds like a bargain to me.
The data retention time is, in fact, 50 years. If the data turns out to be usable through the years than this will turn out to be invaluable. The film industry currently has a crisis on their hands where more and more "garbage" video of sets when no filming is going on and alternate scenes, interviews and all the things we see on the "extra features" sections on those fancy new blu-ray and DVD discs. They need some way to safely and easily store that media for many, MANY years before the common media supports it, and it the discs last 50+ years? This will be a boon. Especially if damaged discs turn out to be as easily recovered as is theoretically possible. The only x-factor will be whether the discs from the first generation of reader/writers is compatible with future generations. If they are? This is a winner.
Their main selling point is longevity. You can store the data on a disk and read it back 50 years from now. Will this company even exist 50 years from now? Will anyone have the equipment to read one of these disks in 50 years? Have they published the specs so you can construct your own equipment, should it become necessary? I don't see this working out. Archiving needs to be done with well-known open standards. InPhase doesn't seem to be off to a very good start in that respect.
LTO-4 tape:
A drive is $5000, and an 800 GB tape is $120. Magnetic tape has a very long, provable, verified and *good* track record at being able to retain data. I've read 30 year old 9-track reels, and have cassettes from the 70's that'll still play.
Their drive is 3x the price, and their media is 50% more expensive for half the space. Their only benefit is the holographic media is random access. Bah. If it's for archiving, who cares about random access?
This gadget smells like fail. Their *only* niche is providing a long term archiving solution with random access, that can't be modified once written (TFA mentions nothing about rewritability). Maybe nice for government or accountability work, but that's all I can think of.
InPhase Technologies announces it will also be releasing Duke Nukem Forever by the end of the year.
When this gets down to $100/terabyte, I will very very seriously consider getting one for my personal use storing my ripped DVD and bluray movies.... Right now I am looking at about 5-10K for a shoestring array to store my DVD and BR collection on for instant access.
Chuck
Nothing in this universe is unlimited.
From TFA:
Which gets us to InPhase's target market: archiving. That's why they were showing at NAB.
I don't get it. No matter how valuable your content, why would you pay $18,000 for a burner and $180 for for a 300GB disc? Just for the price of the media, you can mirror your data across three different brand-new hard disks. Surely the odds of 3 hard disks failing at the same time are lower than that of an untested, brand-new technology with no redundancy?
Maybe I'm too thick, but why would anyone buy this at this price? (Other than the coolness/my dick is bigger than yours factor, of course.)
I see a lot of short-minded folks here, who are completely missing the point of this. The technology has just debuted and you are already putting it down. I bet you are the same people who hate Vista right now, and who are using XP, but hated it before it became mainstream, same with Windows ME, etc.
- We were also told CD and DVD storage was long lived. While 30 years can be expected of a few of the highest grade disks http://club.cdfreaks.com/f33/taiyo-yuden-faq-178622/ 3 years is what most of them manage. Theoretical limits typically don't make it past manufacturers.
- It may indeed last 50 years, but will the equipment it's to be connected to? I've got the first 100MB drive to hit the market. It has lots of stuff on it I want to retrieve. It's a good thing I've kept the 18 year old Apple IIgs it's inside of operating.
Better implemented on solid state holographic storage, but still possible on disk, is the reverse processing of image to beams. (There's a SciAm article from 1995 or so on holographic storage, particularly solid state, that covers this).
Store lots of images on the disk. Illuminate it with a hologram of a target image. Out of each image comes copies of the original reference beams, at a strength proportional to the similarity of the stored image to the target image. Nearly instantaneous, simultaneous retrieval with correlation score built into the signal strength. Lost is the different angles that'd be had in a solid state device, so scanning the disk for reading all the beams and finding those of interest might take a bit longer. The entire US government fingerprint files could fit on one disk and the whole thing searched in seconds, as is often seen on TV. Using it for movie storage makes marketing sense, especially with the initial price tag of $18,000 and disks being $180. But leaving it at that would be a damn shame.
"I may be synthetic, but I'm not stupid." -- Bishop 341-B
You are totally right about most other options not being reliable, the worst being HDDs; you need to keep those spinning regularly and replace them every few years.
People are still reading their decades old MO discs that have been left on the shelf.
Lots of organizations have the need to archive their data and currently the only game in town are MO ($10/GB) and UDO (slightly less) with drives costing $3000+.
That makes the TCO of an 18K drive with 50c/GB very, very attractive to this market and that is what this system is competing with.
Now the only thing that needs to happen is for the technology to be licensed to other players because most CIOs are unlikely to put all their eggs in one basket.
Why is this thing based on rotating media? I thought the point of holograms was the interaction of the 2 lasers and the image/data can be read from any angle...
I wonder if the porn we "archive" on it will show up in 3d if we open it up and view the hologram... see it take a life of its own by giving it so much data in a small area :)
Think about those early 10 megabyte hard drives. Take that form factor and blow it up over the same length of time and you get some crazy-huge numbers. A third dimension to play with? That's like going from DC to AC in terms of complexity and possibility. Interestingly enough, the establishment resisted AC as well. I half suspect that the math simply demanded more brain power than the old school engineers were willing or able to invest.
I remember the day when a roommate took the indoor cat out to the roof. The cat saw the sky for the first time and wet itself, flattened right to the ground and was basically reduced to a form of catatonia. After living in a one-floor apartment, (two-dimensional), being presented with a whole lot of up and down created a great deal of irritation.
-FL
it's coming pre-loaded with Duke Nukem Forever.
I remember hearing about the 4.6GB of storage back in the mid-90s, and it was quite underwhelming when it came out. MO never really took off, long-term. This, I think, may be different.
If the technology in this stuff pans out and can be developed economically and scale well over time (MO didn't), I think it has some real opportunities to take off in certain sectors. It's not for everyone, but neither are rackmounted RAIDs, iSCSI and tape loaders.
For naysayers: do any of you think that this company WANTS to release a boat anchor device like it seems to be going by their pictures? If what the company says is true, and this is not vaporware, the physical size of the drive may be a worthwhile trade-off in terms of capacity and reliability. As technology is developed, processes shrink, things get cheaper, and storage capacity gets bigger. I remember old MO drives being big, and as some pointed, out, a single CD-R costing $40.
I'm not going to buy this thing, but I'll certainly be watching its development in the marketplace. It's interesting to watch, just like I did the Apex back in the day.
Which means there is no going back once the disk is written.
Just thought I'd mention it.
http://www.inphase-technologies.com/products/default.asp?tnn=3
Its got moving parts. Wake me up when its true holographic storage with no moving parts.
"The linear storage we've seen to date has been like the Formula One race track of development, and people have come up with some very clever techniques to squeeze every scrap of use out of it, but really, we've been locked in two dimensions for all this time. Adding a third dimension is watershed stuff."
Whoa, whoa -- slow down, egghead!
If I have seen further it is by stealing the Intellectual Property of giants.
Can you stick these media on the hull of a ship, sail it around the world and then read the data?
"I don't appreciate your analogy very much, Tesla fanboy"
Give me liberty or give me kill -s 9
This has got to be about the sixth time this company has gotten astroturfed onto Slashdot, always with a product "just around the corner." I'll believe it when I see one on a data floor.
In any case, it is a whole big pile of useless. Let's go over the flaws, keeping in mind the alleged target market for long-term archive storage. (As opposed to their last "target market" of near-line storage. Since it's only WORM, that kind of shot that down. Whoops!) In this market, the competition is tape, NOT hard drives. (Hard drives suck for long-term archival, for reasons mostly having to do with fragility and stuck bearings.)
1) The only product even promised for sale in the near future has a puny 20MB/sec data rate. We are talking five years or so behind tape in this regard. Given their schedule track record, don't hold your breath waiting for the later generations.
2) This is a compnay with a ZERO track record of shipping product. Why on earth should we believe their longevity claims?
3) How does this compete at all with old-fashioned analog storage? If I really want to archive visual media until the end of time, nothing beats old fashioned film. It degrades gracefully and has a century-long track record. (Not that that track record is universally good, but it has much improved over the years, and is well-understood.) Long-term storage of audio is trivial, as pressed CD's don't have the long-term issues that CD-R's do.
4) The capacity is a joke. 300GB? Is that the best they can do? How long have they been promising this? Current LTO is 800GB, and it actully is shipping, from companies that have a long history of generally knowing what they are doing.
Whoops! Looking at their "Markets" page, looks like they haven't given up the long-term data storage market! Yeah, they are not going to sell a single one to that market. Why?
1) They mistakenly assume that companies actually WANT to read data a decade down the road for regulatory compliance purposes... They don't! They just want the regulator to be satisfied they made a good-faith effort to do so.
2) Not doing compression on the drive makes it a non-starter here. Business records are VERY compressible. Enterprise tape drives have always compressed data. Not doing so either kills CPU or sucks media.
3) Check out that chart showing the "advantages" of this crap. Let's pick out the lies:
A) "Capacity Roadmap" - The LTO roadmap goes considerably beyond the next capacity bump to 1.6TB, and it is at 800GB NOW. The InPhase roadmap stops at 1.6TB, and they are only at a puny 300GB right now.
B) "Transfer Rate Roadmap" - Pretty much the same problem. They are comparing their theoretical roadmap with what tape is shipping today.
C) "Media Archive Life" - 50 yrs? Based on what? LTO's spec is 30 years (not a measly 10), which isn't bad, and I am more likely to believe LTO's promises than these guys.
D) "Media Price" - Tape is $.25 - $1.00 a Gig? Um no. Not even close. A Gen 4 cartridge can be bought for about $.10 - $.12 a Gig. Their prices, on the other hand, are at the max end of their range. (How can a company with only one product have a range of media prices? The mind boggles.)
E) "Media Handling Issues" - While I suppose this could be a theoretical advantage for offices, it's not for studios. They already have oodles of low temp and humidity storage for their film.
F) "Physical WORM" - Businesses don't care. They are just trying to satisfy regulators, for which WORM on the drive is good enough. This is just a nice way for InPhase to say that their media is not reusable.
G) "Random Access" - What a joke. If your target market is long-term archival, why on earth would you need that?
H) "HW Security Features" - Err... LTO tape encryption? Did they not get that memo? What on earth is "Optical Encryption" anyway? Unless there is some kind of key exchange with the host, that is not encryption at all.
4) They talk about using it for near-line storage, but it is a WORM technology rig
It's an improvement for digital storage mediums for sure, but still pales in comparison to film-based archiving. Microfilm has a life expectancy of 500 years under proper storage conditions. A film-based archival master backed by multiple digital archival master and "use" copies is still the way to go.
It's really embarassingly bad. It makes 2 claims:
1. A small fragment of a hologram can reconstruct the entire data image. The fragment won't let you move as far around the image, but for 2D images, like a photograph, it means a scratch isn't fatal.
This is complete nonsense. A fragment provides a *reduced quality* duplicate of the data image. This is not so bad for photographs, but for digital data it's critical. Bit basic information theory says you can't recover the full image without actually storing the full image.
2. Data density is theoretically unlimited. By varying the angle between the reference and illumination beams - or the angle of the media - hundreds of holograms can be stored in the same physical area.
Again, complete horse hockey pucks. Storage of additional images on a physical medium is certainly possible, but the ability to control the aforesaid 'angle' and recover meaningful data is not infinite. It's limited by the theoretical factors like optical diffraction and resolution, and by the spatial resolution of normal matter made up of real molecules.
Looks like I can say hello to a new format way, skip out an Blu-Ray, and look forward to my super-extra-dooper HD HD TV, with accompanying storage, where I still can't tell much of a difference.
Truth is, this would probably be a sweet format for distributing digital movies to cinemas.
Jumpstart the tartan drive.
If they can't get the price down to the level of a consumer product, then yes, it's doomed.
Tape has become too expensive for the consumer market, though. I don't know if this was the cause of or a result of the increasing concentration of the tape industry during the '90s, but the result is that the only credible backup media for the consumer is hard disks. If their media is $180 for 300 GB it's cheaper to buy a disk drive than a cartridge.
And that goes for tape, too. I struggled with home tape backup for years, and finally just pulled in my last set of archives, burned them to DVDs and kept a copy in external hard drives on the shelf. Tape drives just became too unreliable, and too unrepairable with the resources of a home user, and too expensive to replace when they inevitably broke.
If they (or anyone) could ship even 100GB media at a consumer price, THAT would be a breakthrough.
Now I back up my copy of Duke Nukum Forever!!
You missed one, dude:
3) No degradation while reading. While this may be true of familiar types of holograms, it is most certainly not true of read/write holographic materials. In fact, degradation of signal during read is one of the biggest problems that held this technology back for so many years.
Good point. That's a subtlety of the particular technology, though. It's not a basic violation of the laws of physics to claim it's true for this approach, unlike the claims of infinite storage and no data loss from corrupting part of the media.
2 TB is 2 lifetimes, if by "lifetime" you mean "month" Let's do some calculation:
Assuming (for simplicity) that each porn DVD takes up an average of 4 GB of data (out of the 4.7 GB available), and assuming all porn DVD titles are single-disc (no double-disc special editions), we have:
2 TB on card / 4 GB per porn title = 500 titles per card.
The only source I have for amount of adult DVD releases per year is here on Wikipedia, so Wikiality applies; but it states 11300 releases/year in the USA in 2002, so let's run with that number.
11300 rel per yr / 12 months ~= 941 2/3 releases per month, so we'll round that down to 941.
500 titles possible on a card / 941 releases in a month ~= 53.135 % of a month's worth of porn vids on a card.
In other words, your "lifetime" is roughly two weeks...one week on standard resolution.
And that's not even counting NNTP or P2P.
Now that I've done the adult film industry a service, can I get laid now?
It's limited by the theoretical factors like optical diffraction and resolution, and by the spatial resolution of normal matter made up of real molecules.
Stability of the beam's wavelength and a long coherence length for both the writing and reading lasers also need to be there for good holography results. Today's compact diode and small DPSS lasers pretty suck pretty hard on both criteria.
I thought IBM would be the first to market with this technology. I remember reading a report from them last century (can't find it now) but I did find this from 1999.
In recordable CDs and DVDs that cause them to have such short life expectancies. Its the organic dyes that are burned by the recording laser. These break down, not the polycarbonate substrate.
Think of how many extras a movie could include. Like the entire collection of filming without any edits as well as all personal recording compiled by stage hands.
"G) "Random Access" - What a joke. If your target market is long-term archival, why on earth would you need that?" Any freaking time I feel like accessing that data is random access. Jusy about *ANY* storage medium is random access, like VHS - I leave it on my shelf until I get the random urge to watch that video, then I go get it and access it.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
So you're saying that people who buy this holo stuff will want to throw it out and replace is with hard disks?
Engineering is the art of compromise.
Yes, any time you want to access the data is random access... But, is a 30-60 second delay that it takes to spool a tape vs. however long it takes to align and spool this beast worth the disadvantages?
"Long-term archival" is a time measured in years. For most users, a few extra seconds to access data they haven't had to look at in a year or two is not a big deal.
SirWired
What it basically comes down to is that they've built a machine that is rated to store 300GB for the astounding price of $18,000. They say that the storage limits are essentially unlimited, but the fact is that this is a 300GB machine, with an unproven shelf life (despite what they say in the article). The math doesn't add up here.
Who is this delectable creature with an insatiable love of the dead?
To the people peddling this crap: you can theoretically get unlimited data density on ANY analogue signal! Digital is for chumps! And those that don't have their head up their...
I'll pick this one;
G) "Random Access" - What a joke. If your target market is long-term archival, why on earth would you need that?
I adminsiter a system right now using TSM, ACS and CommonStore that fits exaclty this profile.
Long term 7+year retention of data that generates a consistently high rate of retrieval operations through the day. They would love being able to random access the data. They do already in fact, on an optical WORM drives. For this particular company and this particular scenario, this holographic media seems perfect. Ok, it needs to be a lot cheaper... but that will change and their requirements will be the same.
Darwin Hawking Blackmore
Only if you have zero noise. But that's like postulating a spherical cow of uniform density; useful for theoretical discussions at the water cooler, but not very useful for real world storage systems.
If J.K.R wrote Windows: Puteulanus fenestra mortalis!
You forgot
;-)
5) Notice when this InPhase "real soon now"(tm) holographic drive story has shown up on slashdot:
2005: April 19th
2006: March 29th
2007: May 18th
2008: April 20th
I think they forgot to run ntpclient on the machine with the April 1st press release cron-job.
If J.K.R wrote Windows: Puteulanus fenestra mortalis!
Unfortunately, chronological erasures will not work. You'll have to feed it a tapeworm to hunt down and destroy any undesired files.
Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
Personally I think the only reason you were modded up is because you mentioned a catatonic cat :-)