Slashdot Mirror
Disk Storage Limits Loom 3-5 Years From Now
evanbd writes: "According to this article the major disk manufacturers all finally agreed on something. And it's not a good thing. Specifically, they all think that in about 3-5 years, the superparamagnetic limit will kick in and current technology will stop getting better. But wait, there's more. New technology won't be ready for something like ten years. I know superparamagnetism has been discussed before, but I hadn't seen it as quite this bad and this much of a sure thing." No matter how quickly storage advances arrive, there are certain dead ends will inevitably appear.
First of all, our storage needs are pretty flexible. If we had ten times as much space, we'd stop using mp3/ogg and switch to a lossless compression scheme. And as digital cameras get better, photos will take up more space.
But the real problem isn't that no device will be able to hold my data. The problem is that such a device will be much larger than I want. They already are. Right now I need about 34GB (most of which is music). Ideally, I wouldn't need a big chunk of metal under my desk to store it. I'd much prefer to have everything on my laptop or even my iPAQ. If I could fit 100GB on a microdrive, I would use it up tomorrow.
While you make a decent point, it's not entirely accurate. Right now IDE ATA-100 drives can theoretically peak up to 100MB/sec. Admittedly this doesn't happen often, but in sequential trasnfers it's common for them to sustain 40MB/sec or so. For non sequential transfers seek time will kill you more than throughput anyway. But you're right that this will ultimately be a bigger problem then it is now.
> Well, the average user just isn't filling up an 80 gig drive.
[...]
> I don't think much more will be needed until more people start using their computers for video recording
That the whole crux of the matter. Most people that predict that our storage requirements will taper off in the near future apply obsolete usage models. For application and non-multimedia data storage even a 30GB drive will last a long, long time. But as soon as audio and in particular video enter the scene, nothing is too much.
You seem to think that the main use of multimedia will be on a traditional PC, which couldn't be further from the truth. The future of multimedia computer equipment--in particular storage--belongs to the embedded/set-top market. We've seen the TiVo and the UltimateTV, but those are just the tip of a giant iceberg. Once their costs come down to the $100-$200 range, they will be as common-place as VCRs. Once manufacturers start incorporating PVRs in TVs and cable boxes, the appetite of the market for storage will explode. Just wait 5 years or so until HDTV becomes more mainstream and the same progams require a multiple of the storage space of today.
Another trend I think you'll see is home media servers. Essentially network-attached storage on which TiVo's, MP3 players and all the other new and wonderful toys of the near future are going to deposit their trash. Once they become plug-and-play, and keeping that re-run of Seinfeld around for all eternity is just a button push away, people will want to store more and more media garbage. As those file servers keep piling up in the living room entertainment center, that terabyte won't seem that large anymore.
The difference being that up until this point, the "XX size is always good enough" argument hasn't been able to be made in reference to the fact that generally standard sizes used today are becoming able to push the output/input limits of our senses on the hardware. For example, the human eye only can distinguish certain refresh rate and a certain resolution in pixel and color size.. a screen with this resolution displaying at this refresh rate (with compression perhaps) is a static number of bytes (don't ask me how many.) Once you have storage space/memory space approaching this size, the question "We'll never need more" takes on a bit more validity since the additional storage will not be used for presentation of stuff but only for internal conputation (ie, a 3d universe has a ton of internal information.) Even in a huge 3d game, the amount of space used for the media of presentation will never go beyond a certain resolution (visually or audiably) both in space and time, since our senses can't make the distinction.
Not that I'm saying we're quite there yet, but we are getting close enough to begin thinking about this realistically.
--
Microsoft solved that years ago...
It's called doublespace....
OW!, Stop hitting me!
Do not look at laser with remaining good eye.
superparamagnetism?
Are you troubled by spooks, spectres or ghosts?
Have you had an out-of-body experience?
We are ready to believe you!
Who ya gonna call?
GHOSTBUSTERS!
--
--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
I can.
Of course I work for a major high energy physicis lab, so YMMV...
--
Poliglut
Except with the current techology, since it is mechanical the bigger the platters, the slower the drivers become. Due to more torque need to spin the platters and more effort to make sure the platters spin in a blanced settings. Making them physically bigger fights against speed. You can have a bigger drive, but it will be slower, not much of an improvement.
Given that I/O speed is still such a huge bottleneck for many operations, and given that you can already have 240 gigs of storage for under a grand, is this really such a bad thing?
Gee, that's a whole new twist on "2010: A Space Odyssey."
These limits are high enough that we'll still need the exponential growth for quite awhile before the limit is hit... 1200 dpi laser printers are pretty much the limit for humans (although 600 dpi is probably enough... it's better than I can detect... I can see the pixels in 300dpi printout though)... at 600dpi, a wide aspect ratio monitor (to be easy, says its 16x9 inches) will be about 52M pixels... and each pixel, at 3 bytes (24 bit color), brings the memory usage per frame to 150MB... currently, movies are at 24 fps, so thats 3.7GB per second of uncompressed video... and, of course, the frame rate can go a lot higher (any good quake player will tell you that the game is noticeably better if the frame rate never drops below 60, rather than 40)...
if you were to store this quality video uncompressed, suddenly a petabyte isn't even enough for a whole movie! and even if you had 1000:1 compression, it's still not large enough that you will never fill it up..
--
Patrick Doyle
I mod down every jackass who puts his moderation policy in his sig. Oh, wait a sec....
ex-pe-al-i-do-cious...
(Apologies to Julie Andrews...)
Honestly, though, who needs a petabyte of space?
You, sir, have obviously not tried to install the latest betas of Windows XP.
--
Those who would give up liberty in exchange for security and DRM should switch to Microsoft Palladium!
...that we'll run out of space. /. article about that.
;-) movies. Never mind the fact that DVD burners are coming down in price to provide us with an acceptably-sized external storage medium.
Look! We will have 200GB drive soon, right, there was a
We'll have 400GB drives in 3 years, maybe a nice round number like 512GB will be the upper limit. Maybe 256GB.
But then you think... I can have a PHYSICALLY BIGGER hard drive! A full-height 5 1/4 one, like the old IBM PC ones, but with superdense and 10000RPM spinning platters.
So that could bring me up to maybe 512GB, or more.
Then I could have an even bigger one, two-times the height, with a terabyte on it.
And who the fuck can fill up a terabyte in an appreciable amount of time? I know I could fill up 100GB in a week or so but a terabyte? It would take a long time to find enough pr0n and warez to fill up all that space.
Relax, people, we're coming to the point at which things are going to slow down a bit. Compression technology is getting more advanced, and there's plenty of room an a 1TB hard drive for some 1GB DivX
There's no crisis here. Companies will still sell hard drives. The spice will still flow.
Specifically, they all think that in about 3-5 years, the superparamagnetic limit will kick in and current technology will stop getting better
Clarification: It's not that hard drives won't keep getting bigger and bigger. It's just that you'll have problems fitting more data into less space. The result of this is that when the 1 TB hard drives come out, they will of course be larger in size than the 80 GB hard drives of today, despite the advances in technology that will no doubt be available by then.
---
I pledge allegiance to the flag...
of the Corporate States of America...
So in a way, this will be good becauase it will make people program without the potential of infinite expansion.
Yes the guy is clueless.
Check out the Vinny the Vampire comic strip
"It is a greater offense to steal men's labor, than their clothes"
Because while the HD capacity doubles every time and then, take a look how much the drive throughput developed in the last decade - while we have Ultra160 SCSI or UltraATA 100, the drives don't deliver more then some crawling 20MB/sec, anyway. And they have been at these speeds (orders of magnitude) for a long time.
So - 1 TB of data is rather problematical to use, without heavy indexing - and even then. Because the biggest bottleneck right now is not the processor, not the HD capacity - but the system throughput (IO,motherboard,RAM etc.)
A number of storage manufacturers have been working on holographic storage along with DARPA and many universities. DARPA has the the Holographic Data Storage System (HDSS) consortium and the PhotoRefractive Information Storage Materials (PRISM) consortium, and IBM, Rockwell and another companies have been working on this. Here is a Scientific American blurb about it:
/
http://sciam.com/2000/0500issue/0500toigbox5.html
Also here is IBM's page about it:
http://www.almaden.ibm.com/st/projects/holography
bbh
Well, the average user just isn't filling up an 80 gig drive. They don't want 200gig drives. So even if we hit a brick wall in terms of storage per inch, I just don't think the commodity market will be harmed. Once 256gig drives and the like are avaliable to end users, I don't think much more will be needed until more people start using their computers for video recording and other high storage requirements. And if the embedded device manufacturers have their way, that will never happen. Remember the niche markets like video editing don't drive the commodity market.
Maybe an upper limit on drives will just decrease the number of models a retailer has to sale and thus increase profit. Plus maybe this will give researchers more time to focus on speeding up existing drives and improving reliability.
Uninnovate - Only the finest in engineering.
IBM's Millipede project is supposed to be viable within the next 5-10 years. So who cares about the clunky old magnetic technology anymore?
Necessity is the mother of invention. Limits are there. We will eventually reach them
But we will advance, just like we have before. I remember the over-used phrase '28k oughta be enough for anybody.' Nowadays your considered insane to try to run ANYTHING on system with less then 48-64mb. Your considered legacy hardware. As the need advances the technology will. The more IN DEMAND something is, the more R&D dollars that will be spent on it, the more important it will become. There are alternatives out there, and we will eventually find them. We just have to keep looking.
I'm personally surprised that we are still using magnetic storage. A few years ago people were predicting that OPTICAL would be the way to go. But apparently it wasn't. Magnetic devices prices got driven down as the technology became cheaper and more in-demand.
Who knows where we may be five to ten years from now, perhaps everything will be stored at a remote location for most people and it will be accessed via thin-clients. Perhaps everyone will have their own miniature raid-aray. Perhaps everything will be stored on miniature removable media, but applications will be served from the net. It will be interesting to find out. But I don't think that we need to worry, as long as we continue to inovate, we will find a solution.
[Something witty and intelligent should have appeared here.]
[Something witty and intelligent should have appeared here.]
{Traicovn}
Your comment gave me a wonderful burst of insight, which I will state as follows:
Dasunt's Law - Everything inside a computer will evolve to a state where it requires active cooling.
1st Corollary - Computers will replace furnaces as a source of heating in the home.
2nd Corollary - Within ten years, computers will come with air conditioning, and will require ducting to the outside.
3rd Corollary - The richest man in the world in the next century won't make his fortune from software, but from the sale of electricity.
The result of this is that when the 1 TB hard drives come out, they will of course be larger in size than the 80 GB hard drives of today
The way I see it, only laptops and other small physical footprint devices will be affected by this. For the rest of us, this will just mean the ressurrection of the 5.25" Full Height drive bay (Remember those?).
The article, as far as I read it, makes no mention on what's most important to most drive purchasers: the price. Most computer users out there look for storage solutions that will fit within their budget, not just their computer case.
So long as the $/MB ratio keeps dropping even after the physical size restriction is reached, I'll still be pretty happy. And I can see no reason why it shouldn't. Once the maximum paramagnetically allowed data density has been reached, the only venue for drive companies to compete with each other will be drive price.
I also suspect the drive manufactures will start concentrating on making their drives faster as another field of competition. So instead of seeing more and more sodding enormous drives, we'll start seeing merely huge drives that are either really really cheap or really really fast. And that's just fine by me.
There aren't any advances to be made in optical drives. We reached the diffraction limit years ago. The diffraction limit comes long before the superparamagnetic limit. You could do near- field optics to beat that, but that has proven to be difficult mechanically because of the close proximity required between head and media.