100x Faster Hard Drive In Lab

Gary lets us know about research out of the Netherlands that has succeeded in reading and writing a hard disk using polarized laser light. The researchers claim this offers a 100-times speedup over reading/writing using magnets. People have been trying for years to write data using polarized light; the secret of the current work's success lies in its disk's materials — gadolinium, iron, and cobalt. Working prototype drives should be available within a decade.

Comments

A decade?

[ScrewMaster]

Working prototype drives should be available within a decade.

Spare me. I've been hearing about incredibly dense optical storage for thirty years now. I have yet to see it.

Re:A decade?

[ILuvRamen]

oooookay, why are you still waiting?
http://www.inphase-technologies.com/products/defau lt.asp?tnn=3
It's been commercially available for years and years and years. 1.6 TB on one disk with 120 MB/sec read speed. Yeah it's write once but still.

Re:A decade?

[janrinok]

Haven't you seen the developments in CDs and DVDs during that last 30 years? Everybody else has! A DVD is an incredible amount of storage when compared to the 5MB (yes MB!) hard drive or even my cassette tapes that I was using in the late 70s.

Re:A decade?

[Fizzl]

So... We had DVD 30 years back?
I bet 70's punch card jockeys would have deemed DVD quite something.

Re:A decade?

[janrinok]

Did you notice that they weren't promising greater storage density, just faster access speeds? This is do-able and still worthwhile.

Re:A decade?

[Kjella]

Well, if you said 30 years ago you could have 50GB on a dual-layer Blu-Ray disk that'd be "incredibly dense". But magnetic media has been moving even faster than that, and now non-volatile RAM looks to be making great strides. Size is not a big issue as they already have 64GB 1.8" drives - which means many hundred GB in a 3.5" form factor, if only they can bring the price down. Speed issues can presumably be solved by internally RAID'ing together chips as the technology matures. Ten years down the line this had better be TB++ hard discs to matter, because either magnetic or flash or both will have continued to improve greatly.

Re:A decade?

[value_added]

Spare me. I've been hearing about incredibly dense optical storage for thirty years now. I have yet to see it.

I think the article is about magnetic storage.

In laboratory experiments, they used laser light to write data to a magnetic hard drive at very high speeds.

So, same drive, but a new way of writing/reading it.

Re:A decade?

[Odiumjunkie]

> I've been hearing about incredibly dense optical storage for thirty years now. I have yet to see it.

Why is this modded insightful? TFA refers neither to data density nor optical storage.

Re:A decade?

[ScrewMaster]

I don't know why you think 1.6 TB makes a particularly dense storage medium. It doesn't, and I hope that they're working on a serious contender because our civilization's needs for information storage are increasing at a decidedly non-linear rate. Most of the PC-using population doesn't have much use for more processing power right now, but we can all use a bigger hard drive. Hell, the various up-and-coming surveillance societies alone are going to drive the need for more bits as time goes on, and if digital content delivery really takes off 1.6 TB is going to seem puny.

In any event, there have been many promising technologies in the past few decades that have fallen by the wayside. These guys are saying we'll see their ideas commercialized within ten years? YEARS? That means they have, at best, a proof of concept and have a whole lot of research and engineering ahead of them before any products actually ship. Odds are, their approach won't prove commercially viable for one reason or another, and this will just be another scientific footnote.

As it happens, I was referring to the so-called "holostore", which was being billed as the next great thing in optical storage in the early 80's. Well, that was when I first read about it anyway (Scientific American, as I recall) and it was supposed to store data in a three-dimensional crystalline format that would be read and written by multiple laser beams. They were promising that several Libraries of Congress could fit on an inch-high cube and that it would be available "in a few years." Well, it's been over a quarter of a century now. I understand that it's a work in progress ... but I still can't buy one.

So, like I said, I'm still waiting. I want to see cheap desktop storage that's a couple orders of magnitude more than we have now. Then I'll say they've fulfilled their promise. Avery Brooks is waiting for his flying cars ... he can have them. I'm still waiting for some real storage. I have a lot of videos I want to keep.

Re:A decade?

[ASBands]

1 TB Hard Drive

I'm sitting next to two computers right now, both running Ubuntu. One was purchased in 1996, the other's hard drives were purchased three years ago. The one from 1996 has a 16 GB hard drive, which, as I recall, was the biggest Gateway offered at the time. The other has four 320 GB drives on a RAID 5 (960 GB/894 GiB), which, as I recall, was the second largest behind the 500 GB drives at the time. 30 times larger in about 8 years.

Perhaps you've heard of perpindicular recording, which started early last year. Pretty soon it's going to be impossible to get a hard drive that doesn't have this new technology. You can easily argue that the technology can't go anywhere after this, but it does offer a 10x storage density increase, and you know somebody will be cramming more data blocks on a platter soon enough.

You see, the great thing about hard drives is that they're not critical to the operation of your computer. My Myth frontend has a 40 GB hard drive. The backend, located in a different room and accessed through the network, has 8 500 GB drives on a RAID 5. With the ever-increasing speed of networks, putting things somewhere else is getting easier every day. Sun has taken this idea to the next level with Project Blackbox. Another great thing is that if you need more space, it's fairly easy to just add another drive to your contraption - something you really can't do with processor speed or memory (to a certain point - 4 GB per stick is the highest I've seen).

I see your point - we don't want a datacenter in the basement of every home, but we don't NEED a better system of information storage NOW. There are a lot of ideas out there; most will fall through, but we'll get one, eventually, and that one will make all the difference in the world.

Re:A decade?

[Anonymous Coward]

Perhaps you've heard of perpindicular recording, which started early last year. Pretty soon it's going to be impossible to get a hard drive that doesn't have this new technology. You can easily argue that the technology can't go anywhere after this, but it does offer a 10x storage density increase, and you know somebody will be cramming more data blocks on a platter soon enough.

Blasphemy. No mention of perpendicular recording is complete with out a link to this.

Re:A decade?

[Zeio]

Working prototype drives should be available within a decade.

Huge reliable solid state storage will have taken over by then. Samsung has 32GB SSDs now.
( http://www.samsung.com/PressCenter/PressRelease/Pr essRelease.asp?seq=20060523_0000257520 )

Latency is very low, and R/W throughput will increase along with capacity. optical / holographic storage is like ceramics, its always the "future."

Re:A decade?

[ArsonSmith]

of course, substituting moore's law (is there a similar for hard drives?) to hard drives, they'll be 128 times faster in a decade anyway.

Re:A decade?

[BosstonesOwn]

I think the storage idea is way over hyped. What we need is fast storage to speed up systems not more , since like you stated we can just add another.

At this point we definately need to see faster drives. I have a feeling that the next break through will be speed. At this point it really has to , that is one field in which there is good research going on. With solid state drives coming the issue may be resolved , but I like many others feel that solid state only has an edge in laptops for now. Until it can take massive reads and writes it will be dead in the water.

At this point the speed of the hard drives are the limiting factor. Maybe 7200 rpm drives will start disappearing when we as consumers demand faster speeds. Most hdd manufacturers have a scsi market to protect. WD doesn't have a big one so the raptors came about. Sooner or later if we keep metal platters we need to spin em faster. I for one look forward to 10 and 15k rpm drives in house hold systems.

Re:A decade?

[Xeriar]

Most of the PC-using population doesn't have much use for more processing power right now, but we can all use a bigger hard drive.

You must be joking - in fact I was tempted to mod you funny instead of posting. Just about all of my customers, family and friends would love their computers to be even faster, but 80% of them aren't even using 20% of their drives. And not a one of the latter group has balked at the price of an external HD, to say nothing of DVD burning options.

In the mean time, I would still like to play Oblivion faster, and one of the simulations I'm writing is hell on the processor. Data storage, on the other hand, is plentiful, though more RAM or some equivalent would indeed be nice.

Re:A decade?

[ASBands]

Personally, I've remained unimpressed by current technologies for "faster" drives. For example, look at these benchmarks of Hitachi's 1 TB solution, compared with many other drives. The only significant difference between the 1 TB and the 15K raptor is the access latency time, which is half as much in the Raptor. However, just like in physical memory, latency doesn't seem to matter at all, as you can see in the benchmarks of file reading and writing, system bootup times and all the other benchmarks - the performance difference is insignificant at best. Who cares if your system boots up .9 seconds faster? Page 8 has some nice theoretical and entirely irrelevant differences in performance.

No, the true future of hard drives will be in reliability. Perhaps we'll have hard drives with four platters internally RAID1-ed. The spinning drive thing has it's limits in capacity and speed. I doubt we'll be able to go much further than 10 TB per 3.5-inch drive and I doubt we'll be able to spin a drive much faster than 15k RPM. Hell, a manufacturer could internally RAID-0 a drive for a performance boost...I hope somebody is working on that.

Re:A decade?

[ScrewMaster]

Well, the CPU manufacturers would beg to differ with you, and actually while CPU and PC sales have been on the wane, peripheral sales have been increasing. Intel has been diversifying these past few years because, let's face it ... there's no particular incentive for people to upgrade their processors anymore. They're simply fast enough now (oh, well, okay ... if you were talking about Vista users I'd have to take your point but that will level out soon enough.) But you're right, many users (particularly Windows users) could certainly use more RAM. But we weren't discussing RAM.

Most people (and I wasn't talking about people like you and me that play 3D games or write simulations, and truly need more horsepower) need nothing that an 800 Mhz. Pentium II couldn't deliver. That or less, as a matter of fact. I know many people that still using the Windows 98 box they bought ten years ago because it does what they want. Now, since were going with anecdotal evidence (I mean, there's nothing like invoking your family and friends to support your claims) I can say that my mother (who is an artist, and a sculptor and photographer) has little use for a more powerful Mac at this point, but is always complaining that her hard drive is full. DVD burner wasn't convenient enough, so she bought another hard drive.

There are hundreds of millions of CPUS in homes and businesses worldwide that do NOTHING but run a few common applications, programs that would just as well be served by a 386, and in fact were at one time. Do you honestly, in your heart of hearts, believe that Joe User, who surfs the Web, gets his email, plays a few DVDs and a couple of iTunes now and then is craving for something FASTER than his 2 gig Athlon? Give me a break. He very well may want a new scanner, camera, printer or hard disk, particularly if a computer-literate friend installed Azureus on his box and pointed him to a good Torrent site (where do you think all those downloaded TV shows go, anyway?.)

Let us not forget the millions of 100+ Gb personal MP3 collections out there. People need space, and their need for it is only going to expand. So while you may believe that processor cycles are in short supply, but the reality is that, as a society, we're awash in CPU power. That realization is where projects such as SETI@Home got their start.

More to the point, storage is something that businesses want more than they want CPU. That's particularly in this day and age, where the goal seems to be to record everything about everybody, and laws like Sarbanes-Oxley pretty much force them to do just that.

Re:A decade?

[Anonymous Coward]

You couldn't get 16GB 3.5" HDDs in 1996. My 1999 machine came with an 8.4GB drive and that was pretty standard.

Re:A decade?

[pentalive]

Perhaps we'll have hard drives with four platters internally RAID1-ed.

And if one of the four platters goes, there's no way to replace just it. If all four platters are in the same HDA, one going could contaminate the other three and cause them to fail too.

If this is done, please put a big "THIS IS THE FAILED DRIVE" led on the outside

Re:A decade?

[wilsonthecat]

What you're really admitting to is having a 4 x 320gb porno store

Re:A decade?

[Anonymous Coward]

Clueless pessimism always gets modded up, noob.

Re:A decade?

[Xeriar]

While it's true that a late Pentium III / early P4 -should- be sufficient, the truth is most programs suffer so much code bloat that people who get tricked into buying the latest and greatest still need more (or at least another DIMM in their machines). While I do have several customers who specifically request that I fix their machine as is, no matter how old (odd, and I do make them understand what their machine is worth - they just do not want to throw away a genuinely useful piece of equipment and I can't really fault that), this is not the demographic with the 100+ gig mp3 collection (how much is a 750gb Barracuda again?).

And you're right, the peripherals market is exploding. But that doesn't mean capacity is insufficient. I imagine that is in part because of people like me telling friends and customers "No, you don't need a new computer, you just need more RAM and/or an external/additional HD" and that's the end of it. They then shy away from the new computer - but they'd still like it if their machine was more responsive, which would be fine except they insist on running some insane number of applications along with Norton getting its fat fingers into everything. "But I like the cute cursors" ...

But I've yet to meet a residential customer that required more than one external drive or otherwise have the equivalent of two large modern disks. The people that do are generally not my customers, for obvious reasons - but people like us are far from the majority. Most small businesses I've dealt with wanted 2-3, primarily for redundant backup purposes. The large organizations I've dealt with were not concerned about the cost of storing data, but having it immediately and readily accessible at all times. This is an entirely different concern.

I'm not claiming my clientèle is representative, but if I were to spot a universal trend it would be towards accessibility, uptime and response time - far more than cost of storage. If a new, reliable, non-volatile technology has a hundred times the speed and a hundredth the capacity at ten times the price of modern drives of its era, it will shake the industry.

Re:A decade?

[GanjaManja]

So terrible it's brilliant!!!

that was awesome!

Re:A decade?

[eharvill]

Blasphemy. No mention of perpendicular recording is complete with out a link to this. [hitachigst.com] OMFG! Amazing!!! You just made my Saturday night (as sad as that may be).

Re:A decade?

[ILuvRamen]

I don't know why you think 1.6 TB makes a particularly dense storage medium. It doesn't, and I hope that they're working on a serious contender because our civilization's needs for information storage are increasing at a decidedly non-linear rate. Most of the PC-using population doesn't have much use for more processing power right now, but we can all use a bigger hard drive.

I think it's dense cuz that's like 50x better than a DVD and about the same size. If it was a 2 platter hard drive, it would ne 3.2 TB and the largest single hard drive is 1 TB I think. Yeah yeah it's write once, I know that but still. Also 99% of people have like 10% of their 160 GB drives filled up. Only geeky ppl like me have more than half of their 450 GB filled lol. Even multi-family PCs tend to run around 40 GB tops so almost nobody needs bigger capacity like you said. I really, really, really, really, really want a faster boot time and my hard drive is a serious bottleneck. Same with playing video files and generating thumbnails and copying files :( hard drives suck lol

Re:A decade?

[Anonymous Coward]

I can guarantee you if that computer was really bought in 1996, the 16gb hard drive is an aftermarket one added around 1999 or 2000. One of the biggest drives I could find in 1996 for consumer level hard drives was around 1.2GB or so, *tops*.

Re:A decade?

[skogs]

Perpendicular /= Vertical

If the bits were standing on end, it would be called vertical recording. That is a horrible video. Perhaps I'm mistaken...but I thought perpendicular meant instead of going around the disk like this:
- - - - - - - - -
they could go like this:
llllllllllllllllll

sigh.

Nothing like marketing and graphic artist dorks messing up real, honest, true, scientific, and conventional wisdom and methods.

Re:A decade?

[Fweeky]

"latency doesn't seem to matter at all" It might not matter much on your desktop, I can assure you it matters quite a lot on plenty of servers. People don't buy things like this for nothing. It's less important when all your performance-critical data fits in memory, but that's not always feasable.

Also, Raptors are 10kRPM disks, and still trade off density for latency at a level most high performance drives do not.

"Perhaps we'll have hard drives with four platters internally RAID1-ed" This crops up from time to time, and I'm yet to quite grasp the thinking behind it. How exactly do you manage the "Independent" part when you put it all inside a single drive? Oh, I know.. make your "drive" an enclosure for two or more smaller ones. Hmmm...

Re:A decade?

[JaumPaw]

In this case it IS vertical, given that you accept that the disk plater is horizontal. (which is the case for the animation, anyway)
Perpendicular means that the angle between the two objects (this case, the disk and the "bits") is 90 degrees.

Check your facts before you rant.

Re:A decade?

[joto]

And if one of the four platters goes, there's no way to replace just it.

As if we even care anymore. This can be said of just about any consumer product, whether it is a HD in a RAID, a battery in an ipod, or a surface-mounted component on some circuit board in just about any electronic toy. Just toss it away. Economies of scale makes it cheaper to just replace the whole thing. The only ones crying, are the environmentalists.

If all four platters are in the same HDA, one going could contaminate the other three and cause them to fail too.

This happens in normal RAID too, it's just that people running RAID now either has a clue, or think they have a clue, so they don't talk about their miserable failures too loudly. In a pre-made RAID-disk, one sane default (for naïve consumer applications) would be for the disk to revert to read-only once a platter has failed. In professional settings, where the disk is monitored by people with a clue, and people complain louder when they can't save their TPS-reports just because a disk fails, other settings can be used. The operating system can be used to configure this.

But anyway, this will make a disk almost doubly expensive, and it will not halve the failure rate. So I doubt it's worth the money. Periodical backups are a better solution (and could also be built into the drive by adding extra platter(s) and some logic that periodically mirrors stuff at various intervals (in periods with low disk-usage), a log of new sectors written, and a small cheap CPU). The main advantage of periodical backup, is that the wear and tear on the backup-platter will be different from that of the main drive, and thus, they are less likely to fail at the same time.

If this is done, please put a big "THIS IS THE FAILED DRIVE" led on the outside

Why? In a typical home or office PC, there are between 1 and 2 harddrives, which are only replaced when powered off. LEDs won't solve many problems there. The only people it will help, are people with many disks, that hot-swaps them, but are too stupid to keep their disks in an orderdly fashion. Well, ok, perhaps there is a use for this anyway...

Re:A decade?

[skogs]

Hmm...I guess I am proven a buffoon. I saw this years ago and figured it was what it said...perpendicular to what is normally written. Still horizontal, but not longitudinal. Sigh. I am stupid.

Vertical it is.

Re:A decade?

[jshriverWVU]

No offense, but you can get a 1TB drive the size of two standard bricks. I would call that pretty dense. Forget 30 years, even 8 years ago I would be tickled with the tought of having that much data storage in such a small place. On that note I was happy when the 1gig drive came out.

So as things progress, by the time I have kids you'll probably be able to get a 100TB drive that fits in the size of single platter HD nowadays.

Re:A decade?

[TemporalBeing]

Just about all of my customers, family and friends would love their computers to be even faster, but 80% of them aren't even using 20% of their drives. And not a one of the latter group has balked at the price of an external HD, to say nothing of DVD burning options.

Are they also the ones that associate the speed of Internet access with the speed of the computer, and they use dial-up to access the Internet? Seriosly - my wife's grandmother does just that. The "Internet" was slow (since she used dial-up - 56k), so she bought a new computer - and still used dial-up. Only thing it did for her was get rid of the junk - nothing a simple re-install/refresh of the system wouldn't have cleared up.

Do we even have to say it?

[Dirtside]

I think this story is a dupe from, like, 1993.

Seriously, I can't think of an otherwise plausible tech that's been vaporware longer than light- or holography-based data storage. I know there have been working examples for years, and I think there's even a (really, really expensive, very specialized) production version or two, but come on! How hard can this be?

Re:Do we even have to say it?

[janrinok]

No, they are not looking at light- or holography-based storage. They are reading and writing a hard drive using laser. That's not bad from the same people that invented the CD (Phillips). I'm still searching Google to see if the same university provided research support to that earlier achievement. Your question is valid however, I don't know why the promised holographic storage has never been produced in large quantities other than it must be proving to be more difficult that either you or I think it should be.

Re:Do we even have to say it?

[commodoresloat]

Seriously, I can't think of an otherwise plausible tech that's been vaporware longer than light- or holography-based data storage. Duke Nukem Forever?

Re:Do we even have to say it?

[Ice Wewe]

Because, if we had optical data cubes, some insensitive clod would replace your precious [optical data cube] with one that has a novelty fly in it, and then you'd loose all your p0rn-- I mean data. Still, I think the idea of an optical data cube (like that in the Hitchhiker's Guide to the Galaxy) that can store massive amounts of data would be an awesome product, if it's even possible. Now if only I could get the co-ordinates to Magrathea...

Re:Do we even have to say it?

[NeilTheStupidHead]

using polarized laser light Isn't light from a laser, by definition, highly polarized already? Are there degrees of polarization?

Re:Do we even have to say it?

[Dirtside]

Duke Nukem Forever?

I said plausible tech!

Re:Do we even have to say it?

[quibus]

They are reading and writing a hard drive using laser. That's not bad from the same people that invented the CD (Phillips). Um, this work was not done by people from Philips, but people from the Radboud University in Nijmegen, the Netherlands. I even know the guys there, I did my physics Master's research project in the same group, the mentioned Hohlfeld was my mentor :)
Anyway, it's really great stuff they discovered. This is truly new! I wish them a lot of good luck in continuing this research.

Re:Do we even have to say it?

[janrinok]

Er, I meant the same people - as in Dutch. And I believe that there was cooperation between Phillips and various Dutch universities regarding development of the CD which I alluded to in my earlier post, although I couldn't confirm that Radboud Uni was involved. But thank you, I understand your point.

Re:Do we even have to say it?

[scotch]

It depends on the laser, and especially the resonance cavity and the reflective ends of that cavity. But in general I think no.

The typical qualities of a laser, kind of a "definition" if you will are: monochromatic, coherence, directionality and power/brightness. Different lasers will vary even in their degree of fidelity to this list.

Re:Do we even have to say it?

[toddestan]

Seriously, I can't think of an otherwise plausible tech that's been vaporware longer than light- or holography-based data storage.

Storage that uses light to read the data has been around for over 20 years. It's called a CD, maybe you've heard of them?

Hard Disk?

[mcfedr]

Hard Disks are old news...no one is going to be using them in 5 years, let alone 10...flash is so the way forward

Re:Hard Disk?

[Anonymous Coward]

You're just trying to distract us with a stick while you're buying stocks in gadolinium mines.

Re:Hard Disk?

[Anonymous Coward]

You mean trying to distract us with a memory stick?

Re:Hard Disk?

[Spy der Mann]

Actually in 10 years, flash will ALREADY be obsolete. It'll be replaced by phase-change RAM or Nanotube memory.

Re:Hard Disk?

[Odiumjunkie]

> Hard Disks are old news...no one is going to be using them in 5 years, let alone 10...flash is so the way forward

Probably not in the notebook/desktop consumer market, but I can imagine enterprise/research uses for magnetic HDDs where read/write times are less important and $/GB much more so.

That said, if I'm right, laser-based magnetic storage being faster than current tech won't really matter for that kind of scenario.

Re:Hard Disk?

[Anonymous Coward]

Don't think so. Flash is pretty much at its max arial density that it will get, so if you want more bits on a flash chip, you will have to start having a larger physical size.

Flash is also not a stable read/write medium... write the same sector a couple thousand times, and you won't have a sector anymore.

Re:Hard Disk?

[mcfedr]

surely if price per gb is more important so new fancy lazer thing is going to be far more expensive than good old flash memory

Re:Hard Disk?

[vininim]

You mean like ebooks have replaced deadwood and ink books?

Re:Hard Disk?

[rolfwind]

It will happen. The problem is that no one wants to read books on a computer screen. Enter E-ink. It has the best of both worlds.

The problem is that the readers are now $700. When it gets to under $99-$199, it will take off. People, being as mobile as they are today, don't want libraries at home anymore that takes shelves upon shelves away. Plus it would be nice to be able to carry your library with you.

Re:Hard Disk?

[Anonymous Coward]

He just said that you fucking tit.

Re:Hard Disk?

[Khaed]

People, being as mobile as they are today, don't want libraries at home anymore that takes shelves upon shelves away.

Um. False. There may be fewer people, but there remains sizable number who do, in fact, want libraries at home. Something about books, and having lots of them, is very appealing to some people. I don't know of any other form of entertainment where people develop the type of feeling they do when it comes to a book they've read a time or three -- the actual, physical book.

Re:Hard Disk?

[Grimbleton]

I'm only 21 and I've already amassed enough books to thoroughly cover my average-sized living room's walls using standard 6' tall 5-shelf bookshelves. The library I'll have when I have my own home, and not just an apartment, will probably rival a small town public library. And I still won't feel I have enough, or have read enough.

Re:Hard Disk?

[Anonymous Coward]

Thanks for the wikipedia link!

In a website like slashdot, most people aren't familiar with complex tech jargon.

Re:Hard Disk?

[Anonymous Coward]

Will another tech replace flash? Seems likely; flash has inherent problems with write-latency and max number of writes. But PRAM or nanotubes within ten years? Are there even any commercial products based on these yet?

I doubt currently non-existent hardware can overtake where flash will be in ten years time on *all* of speed, price and capacity. Flash is still improving exponentially. Wiki claims NAND flash is 18 years old, and it's not yet caught up to harddrives: similar in speed (better latency, worse sustained), 10x worse in capacity, 10x or more worse on cost.

However, I would love to be wrong - I am eagerly waiting for SSD prices to halve (maybe next year?) before replacing my OS drive. So another order of magnitude performance gain past that in only ten years time would be great :) 100 gig of ram-equivalent per pc? That's some fun home supercomputing possibilities!

Re:Hard Disk?

[Anonymous Coward]

s/thousand/million/

Re:Hard Disk?

[MrCrassic]

I highly disagree. So far, it seems that the fastest and most commercially available form of long-term storage technology right now is the hard drive. Flash media is up and coming, but the write operation for these devices are horrifically slow for any sort of hard disk replacement. The closest new breakthrough that will "replace" the hard disk is a "hybrid" hard disk, which is still a hard disk anyway.

If I were a stock investor, however, I would definitely put my money on these new RAM technologies coming up, as they do look promising for the years ahead.

Re:Hard Disk?

[Zak3056]

Probably not in the notebook/desktop consumer market, but I can imagine enterprise/research uses for magnetic HDDs where read/write times are less important and $/GB much more so.

If cost/GB were more important than performance, the 5400rpm IDE hard drive would be king of the data center.

Re:Hard Disk?

[Willuknight]

and thats why you buy a cheap PDA (i got one for about $30USD, and use it to read books.

I started off with a palm m500, and now have a $300 htc harrier / pda2k.

I have read so many ebooks on my pdas that i've lost count.

Re:Hard Disk?

[rolfwind]

Well, I think you missed my point. I did not mean they don't want books, I meant they don't want the downside of books. The space it takes. Americans move an average of every 7 years. In Europe, less so, but there you have to contend with apartment size, same with Japan. I would imagine, that if people were offered to store all their reading in a space the size and weight of a loaded shelf or one the size of an ereader + external harddrive - most would take the harddrive.

I had to let go of a lot of books simply because of space. Same with newspapers and magazines. I am sure most people are like that, at least with newspapers/magazines. Once a solution is offered where all this can be kept locally but in a convenient and nonintrusive way, I don't see why paper is preferable to e-ink.

As for actual physical forms of books, e-ink emulates the page. It does not have blacklight and actually looks like a printed page. Perhaps flipping through the book won't be as effortless -especially on the first models, but there will be upsides such as instant search, that paper could never match.

Ebooks just never took off before because of the screens, who wants to stare at a backlighted, computer screen all day? That's not relaxing.

If you really meant that they want loads and loads of bound paper, well, I love to read, but I could do without the hassle of a physically printed page.

Re:Hard Disk?

[rolfwind]

If it works for you that is good, but the screen size is too small and the backlighting kills it for me. That is why e-ink is exciting to me.

Now only if the readers would sink in price. Larger screens (8x10 please) would seal the deal, and color would also be good but not 100% necessary yet.

Re:Hard Disk?

[oakgrove]

As an extensive traveler, avid book consumer and owner of a Sony eReader, I can honestly tell you that it's not going to replace dead tree books any time soon. It was fun at first but, now I hardly ever pick it up. And that's not just from lack of content, I know how to use utorrent. It's just a poor substitute for the real thing. One glaring flaw is I can't flip through the pages, scanning for what I'm looking for in a reference book. That's a real deal killer right there. I also like a bisected form factor in my reading material. I don't want to feel like I'm reading some sort of scroll.

To sum it up, books, not to mention magazines won't be going anywhere anytime soon if this thing is any indication.

Re:Hard Disk?

[djl4570]

I thought the same thing. Flash memory or whatever replaces it will be much farther along in ten years. Dell is already offering a flash memory hard drive for laptops. Low power consumption, no latency, and less heat. Unless this vaporware is faster, better and cheaper it will die in the laboratory. In 2002 I bought a high end digital camera. 128MB flash memory cards were about half the price of 256MB and I think the 512MB had just come out and were very expensive. Five years later 2GB flash memory cards are often on clearance sales for less than half of what I originally paid for my pair of 128's. Non mechanical storage has the potential to change the shape of laptops as well. A flash based system does not have to conform to the form factor of a conventional hard drive. I'm sure they will at least for now for compatibility but someday that form factor constraint will follow ISA into oblivion. There are several companies that make solid state flash based hard drives in both 2.5 and 3.5 inch form factors. I wonder if a server would benefit from a pair of these used for system and paging files.

Re:Hard Disk?

[WuphonsReach]

I'm betting that whichever e-ink book reader manufacturer that provides a seamless interface to Project Gutenberg files will probably take the kitty. Especially if they come up with an open-source / open-spec / no-patent specification for handling user notes, highlighting, and other marking up of the text.

I've been looking at the Sony reader. I think it's probably the closest of the bunch, but I'm not convinced yet (Sony tends to be too proprietary in their thinking).

Re:Hard Disk?

[AbRASiON]

You're kidding yourself, absoloutely kidding yourself.

Everytime flash comes even close, hard disks decrease in cost and tripple in available space.
Please to be showing me the 1TB 90mb write, 60mb read sustained flash drives under 400$ right now.

Ok show me something even close to that, without read / write cycle problems.

It's not going to happen, they will always remain behind, at best we'll get little ones in our machines to hold the OS only or something but not the real data.

Re:Hard Disk?

[Khaed]

I'm betting that whichever e-ink book reader manufacturer that provides a seamless interface to Project Gutenberg files will probably take the kitty.

If you can skip the 20 pages of shit at the beginning, yes.

I really love Project Gutenberg, but the amount of crap some of the books have at the start is just a pain in the ass sometimes.

I also wish we could revert copyright law so that books made since the 30s would be in public domain. I'd like to see someone argue the point that a book published in 1950 has some sort of need to remain protected by copyright law for another seventy years.

Re:Hard Disk?

[toddestan]

If cost/GB were more important than performance, the 5400rpm IDE hard drive would be king of the data center.

If they were still available in high capacities, I would be using them a lot. Cooler, quieter, use less energy, cheaper, and generally more reliable. Many mass storage applications do not need a lot of speed, 90% of the data on my computer (music, videos) could be moved to a 5400RPM or even a 3600RPM drive with no noticable impact on performance.

Re:Hard Disk?

[rolfwind]

I suspect it will take several iterations to get a good reader, like 2-5 years down the road.

I like the sony, but for my reading (textbooks) I'm hoping a reader with a 8x11 inch screen comes out. That would be a great size. Color could come later as it's nice but not a dealbreaker.

Re:Hard Disk?

[Zak3056]

Many mass storage applications do not need a lot of speed, 90% of the data on my computer (music, videos) could be moved to a 5400RPM or even a 3600RPM drive with no noticable impact on performance.

Note, I said "king of the data center." I agree that the devices are adequate for your application. I disagree with the grandparent's suggestion that price/GB is more important than performance in enterprise computing.

Re:Hard Disk?

[toddestan]

While I'm sure there are high throughput servers that absolutely need the fastest disks possible, there are other servers just seem to sit around mostly idle. The FTP server really doesn't need a 10K scsi drive, but it seems to have one anyway. Same with the license server and the alarm server. I could probably replace half the drives in a typical datacenter with 5400RPM drives and have no impact on performance as far as the users are concerned. Though admittedly, the best thing would be to just virtualize all of those relatively low demand servers onto one physical box rather than building a half dozen lower power boxes.

Re:Hard Disk?

[Willuknight]

i actually think i'd prefer a smaller screen to something the size of a book, 2-3 inches is enough, can be carried in your pocket and is easy to read (for me). Whats wrong with backlighting?

link...

[cosmocain]

...to the original publication.

the really fascinating thing is not THAT they succeeded to change the magnet field via lasers, it's the speed if you compare their figures to this

I think I'm dyslexic...

[Anonymous Coward]

Instead of "Netherlands" I read "Neanderthals" and I begun to wonder... since when Neanderthals had lasers !?

Re:I think I'm dyslexic...

[mcpkaaos]

Since January 20, 2001.

Re:I think I'm dyslexic...

[skoaldipper]

Instead of "Netherlands" I read "Neanderthals" and I begun to wonder... since when Neanderthals had lasers !?

Thetans were the original Neanderthals after Xenu booked them onto a DC-8 liner. I believe the Teegeeack Safety Administration (TSA) confiscated them shortly after departure into Krakatoa.

Quite frankly, my concern is running linux 10 years from now with these new laser drives. A lot of applications discard items to the black hole /dev/null. The escape velocity of these photons might just be insufficient, redirecting information to /dev/oldmem instead, thus reawakening lost Thetan memories trapped inside my computer. Surely, as unsuspecting emails on linux beowulf clusters around the globe are automatically sent to helpme@scientology.org, this will usher in the great Auditing Apocalypse we all fear.

Re:I think I'm dyslexic...

[newr00tic]

..They^ve had it ever since the day when they learned to say "Imma fiering mah..[laz0r]," naturally.

Re:I think I'm dyslexic...

[Anonymous Coward]

Instead of "Netherlands" I read "Neanderthals" and I begun to wonder... since when Neanderthals had lasers !?

Lasers - so easy, even a cave-man can do it.

Faster how?

[DaleGlass]

The article is unclear on the details. Are they making a hard disk with an optical head? In that case will it really help that much, given the problems with making the disk spin faster, and the seek latency? There are 15K RPM drives already, only they're a bad idea for consumers as they're noisy and require cooling that's not available in most consumer oriented computer cases.

Re:Faster how?

[janrinok]

I guess you didn't RTFA:

... they used laser light to write data to a magnetic hard drive at very high speeds. The technique works because the photons transmitted by the laser actually carry angular momentum, allowing them to interact with the hard drive. Also, each laser pulse heats a tiny space on the disk just enough to make changing its polarity--thereby storing a bit of data--a little easier. The key is reversing the polarity of the laser pulses, which can produce the equivalent of either a 1 or a 0 of binary code on the disk storage medium.

Re:Faster how?

[DaleGlass]

Yes I did.

Hard disk speed comes from several factors:
Data density: The more densely it's packed, the more data per second passes under the head
Rotational speed: The faster it spins, the more data per second passes under the head
Latency, a combination between the seek latency (how long it takes the disk assembly to move to the location), and rotational latency (how long it takes for the platter to rotate to the required position), determines how long it will will take the disk to start reading data from somewhere else.

They don't explain how the laser is mounted. If the laser is sitting in the same place as the current magnetic head, then that the head can potentially read/write 100 times faster doesn't really matter, when there's no way the disk itself can be made spin 100 times faster. 7200RPM are about the fastest you can stick in a normal case without extra cooling without it melting.

So, the disk can't be made to spin much faster, making the assembly move much faster is difficult and bumps into rotational latency anyway, and they aren't packing data more tightly because they admit the footprint of their laser is bigger than used by current tech. So again, even if their head can read/write 100x faster, does it even matter given that it'll never be given the opportunity of doing so?

Re:Faster how?

[janrinok]

OK.

Are they making a hard disk with an optical head?

Yes. Or at least they are to my interpretation of using a laser to write to the disk. You can be pedantic if you wish but they haven't claimed something that they haven't done.

And no-one is arguing with any of your other points, which I guess is why they reckon it will take a decade to come up with a workable, deployable solution. Perhaps they are being optimistic but, hey, who knows? The world is full of things that once looked impossible but are now taken for granted. The clever part is that they are changing a magnetic medium using a laser. Lets accept that and now address the other problems.

Re:Faster how?

[DaleGlass]

Well, that's just the thing that makes me wonder if there's any point in going in this direction. Spinning that much faster would require some really good bearings and a platter made of unobtainium (IIRC, at the current speeds, the forces trying to shatter the platter are quite significant already).

Besides, it seems that the new way of doing this is with Flash or something similar. I wouldn't be surprised if that's what we'll have everywhere 10 years from now. No seek latency, you can get more speed by internally RAID-ing it, low power usage, no noise, less heat. Flash has a write limit, but unlike hard disks Flash fails predictably and nowhere near as horribly as a hard disk. A Flash drive should be able to very precisely (unlike hard disks even with SMART) tell you it's about to run out of spare sectors.

Re:Faster how?

[janrinok]

Perhaps we are not seeing the potential. Maybe the advantage will not, ultimately, be manufactured in hard drive terms. But I suspect that there will be a good few bright people thinking of ways to use the fact that you can change the field on a magnetic medium using a laser. If the read/write speed is increased and size of the magnetic field on the medium is reduced by an order or 2 of magnitude then perhaps someone will have a bright idea of how to convert the theory into a working, usable device. Perhaps rotating the medium is not the way to go and it might be possible to control the laser using an alternative technology which makes the whole thing feasible. I'll meet you here on /. in 10 years time and we can see where the path has taken us.....

Re:Faster how?

[MattBurke]

If the laser is sitting in the same place as the current magnetic head, then that the head can potentially read/write 100 times faster doesn't really matter, when there's no way the disk itself can be made spin 100 times faster.

But what if they could make the laser deflect (is there some non-moving way of doing this?) in a tiny but precise arc millions of times a second? With multiple receivers that would enable it to read bits from several tracks during the time it would normally be sitting there waiting for the next bit to come along...

Re:Faster how?

[tomz16]

As far as I can tell, the advantage in this approach is the fact that you are using light, which can relatively easily be redirected. This would allow you to overcome the mechanical limitations on rotational speed.

Right now you have to move the writing surface under the head mechanically. In the future, you'll be able to deflect the beam to the proper location on the drive in one or both axes.

Re:Faster how?

[some guy I know]

Right now you have to move the writing surface under the head mechanically. In the future, you'll be able to deflect the beam to the proper location on the drive in one or both axes.

Maybe, but the fact that there isn't any product available now (about which I know, anyway) that does this with a CD/DVD-like medium indicates that there are difficult problems to be overcome in this area.

Stupid hype

[imsabbel]

Actually, this couldnt have less to do with data storage (you cannot really focus your femto-second laser down to spotsizes lower than what we currently have in HDs, plus there is no real way for a femtosecond source that not bulky, wastefull and expensive).

On the other hand is the switching of magnetic domains by the polarity of a circular pulse an archivement in itself. But of course fundamental research doesnt interest anybody, so they have to create a stupid "next storage medium" out of it.

No, it isn't really 100x faster

[geophile]

The researchers managed to transfer data at intervals of about 40 femtoseconds, or quadrillionths of a second, about 100 times faster than conventional magnetic transfers

That optimizes a tiny part of the problem. There are two mechanical issues, 1) waiting for the right part of the disk to rotate under the read/write head, and 2) arm motion. Without eliminating one or both of these delays, I don't see how this leads to faster secondary storage access in practice.

Re:No, it isn't really 100x faster

[janrinok]

Which is probably why they said that it will take a decade to produce usable devices. However, that doesn't detract from the discovery or achievement. It is another hurdle passed which will let someone else concentrate on solving the other problems.

Re:No, it isn't really 100x faster

[Mike McTernan]

Could you not eliminate arm motion with a very fast rotating mirror, constantly passing the laser across the rotating disk? Then you just have the problem of rapidly switching the laser on and off to target the specific track.

Re:No, it isn't really 100x faster

[kasperd]

Without eliminating one or both of these delays, I don't see how this leads to faster secondary storage access in practice.

Those delays may be a bit inconvenient, but they are not a major problem. For more than a decade, we have more or less known how to deal with them. Yes, a bit of research is still happening in this area, mainly for two reasons. We want to do things with disks that we didn't always do, such as virtual memory. The other reason is, that as CPU spees have grown faster than disk speeds, the most efficient file system designs are not the same as a decade ago.

Currently we have a much greater problem than seek times. Disk capacity is growing faster than transfer speed! That means the time to read (or write) a disk from end to end (that means without seeks) will grow with every new generation of disks. A decade ago you could read an entire disk sequentially in about 20 minutes, today it takes hours. During the last decade the transfer speeds have not grown by a factor of 100. A decade ago a new disk could transfer a few MB/s. Today they I have not yet come across a single disk that could do 100MB/s. If the next decade will give us a 100 times improvement in transfer speeds, it is much welcome, because we need it.

Sure it may mean we want to do file systems with larger block sizes. They already have grown from a typical 512 bytes to 4KB now being typical. (At some point 32KB was common, but was so because of shortcommings in the file system rather than performance reasons). Sure we'll have to solve problems such as how to boot an operating system or load an application with as few seeks as possible, but we are slowly getting there already. And once we get computers with 128GB of RAM, it will be nice being able to suspend to disk and resume from disk in less than 20 seconds.

Re:No, it isn't really 100x faster

[jadin]

Is a rotating disc really necessary with optical? I'd imagine (quite probably incorrectly) that you could aim a laser wherever you wanted to without moving anything if designed it right.

Re:No, it isn't really 100x faster

[NeilTheStupidHead]

I doubt it. Given the size of even a standard HD you're going to end up with the laser hitting the surface of the disk at an angle very far from the normal. You might have better luck with an array of lasers offset so that each only has a small range of motion/tilt but covers the whole surface of the disk.

Re:No, it isn't really 100x faster

[nephridium]

I don't think the crucial part is how (well) it works, but that this technology works. It's one of the puzzle pieces that might lead to better mass storage media. Other things will need to be researched as well (as stated in TFA, e.g. increasing the data density), but this definitely sounds promising.

The cool thing about lasers is that the data can be transmitted through the air/vacuum and is not reliant on quasi-physical contact of the reading/writing head; so while the old "arm over spinning disk" might work as well (simply increasing the data density alone would yield greater transfer rates), one could also think of an optical array that bounces off a laser (or even a few simultaneously) with a minimum of mechanical moving parts.

Re:No, it isn't really 100x faster

[ceeam]

You're absolute right but think about bits density and as a result the storage capacity!

Re:No, it isn't really 100x faster

[Stellian]

During the last decade the transfer speeds have not grown by a factor of 100.

I suspect a lot of that can be attributed to the market demand, rather than an actual technological limit. The size of the hard-drive is it's main metric, and the only thing that consumers look at, and of course the engineers will make a compromise size/speed/price.
At the end of the day, what would you rather have:

• A fast, 15.000 RPM, 16-Platter, energy hungry beast, that makes a horrifying sound every time you access a file (because of the large actuators need for small acces time)
• A small and energy efficient 10x larger, 10x slower drive, fast enough for watching porn and browsing Slashdot

And if you really want the first option, you can have it as a SCSI, of course at a price premium since it won't benefit from the economy of scale the slow IDE drive has.

Re:No, it isn't really 100x faster

[Anonymous Coward]

When you increase density, keeping seek and rotational latency the same, the thourgh put increases. For random seeks, the drive speed is only slightly faster, but for sequential reads, the drive is much faster, both due to the speed up of the density, ( and this is the part you may be missing) and the smaller number of seeks that it needs to read a given amount. In the real world ( i.e. booting up ) it will speed up the initial OS loading( sequential reads ), but the extra programs such as virus software and anti-spyware is more of the random seeks.

Drum memory does away with seek time, you just need a head for every track. Multiple heads can cut rotational latency, and both of these technologies have been availble since the 50s. ( early winchester drives and the Drum storage from the 70s on HP2000s ).

Very important I think is to improve the seek scheduling. VMS and later Novell had the best ideas, and that is why Micro$oft hired the guy to do NTFS.

But most important is to make great optimizing software. We are still in the dark ages on this.

Art.

Re:No, it isn't really 100x faster

[geophile]

Sorry if I wasn't clear -- I was responding to the slashdot writeup, not to the researchers (obviously)
or the article.

Re:No, it isn't really 100x faster

[scotch]

You don't even need a mechanical mirror mover - you can direct and refocus light very quickly using solid state tricks with LCDs that modify their refractive index locally.

Re:No, it isn't really 100x faster

[Fweeky]

"I have not yet come across a single disk that could do 100MB/s" Seagate Savvio 2.5" 15kRPM SAS disks can manage roughly 110MB/s in good conditions.

Seagate's next consumer drives, 7200.11, supposedly manage similar serial transfer rates with their 250GB platters, though will of course have massively higher seek times.

Re:No, it isn't really 100x faster

[kasperd]

I suspect a lot of that can be attributed to the market demand, rather than an actual technological limit. The size of the hard-drive is it's main metric, and the only thing that consumers look at

It is not just a question of demand, but also what specs are actually provided. In many cases it doesn't say anywhere what the transfer rate on the disk is, that is something you will only know once you have bought the disk. If the manufacturers wanted to, they could tell what the minimum guaranteed transfer speed was. Eventually customers will have to consider this, because if the current trend continues, at some point the capacity will no longer matter when the transfer speed is too slow to fill the entire disk within its lifetime. Of course that is still far away, too far away for most people to worry.

At the end of the day, what would you rather have

With those options, I'd take an array of IDE disks rather than SCSI any day. Let's for a momemnt consider what are the advantages of that SCSI disk over an IDE disk. Maybe the transfer rate is higher, but it is not twice as high, so an array of IDE disks would still be able to match the transfer rate of the SCSI disk. The SCSI disk may have lower seek times, but like I already said, that is less of a concern for me. A sequential read of the SCSI disk will probably be significantly faster than the IDE disk, but that is mainly because the SCSI disk is going to have a smaller capacity. I really don't see that as an advantage :-)

Re:No, it isn't really 100x faster

[Mike McTernan]

Or you could just use two mirrors, one spinning near the laser source, and a second to deflect the beam down onto the surface.

Re:No, it isn't really 100x faster

[munch117]

There are two mechanical issues, 1) waiting for the right part of the disk to rotate under the read/write head, and 2) arm motion. Without eliminating one or both of these delays, I don't see how this leads to faster secondary storage access in practice.

You're visualising a conventional harddrive design, with a spinning disc and a read/write arm and all. An optical storage device need look nothing like that.

One nice thing about optics is that you aren't limited by the speed by which particles travel through space. You can move the endpoint of a beam of light much faster than that. Even faster than the speed of light.

Re:No, it isn't really 100x faster

[Anonymous Coward]

is it's main metric

"its".

Reverse the polarity of the tachyon pulse!

[jollyreaper]

Gary lets us know about research out of the Netherlands that has succeeded in reading and writing a hard disk using polarized laser light. Oh my god, dicking with the polarity actually did fix something! I take back half of the mean things I've said about Wesley Crusher.

Re:Reverse the polarity of the tachyon pulse!

[Anonymous Coward]

Oh my god, dicking with the polarity actually did fix something! I take back half of the mean things I've said about Wesley Crusher.

Unless it made a red light turn green or a hot plate go cold, he's still on airlock testing detail.

Re:Reverse the polarity of the tachyon pulse!

[ScrewMaster]

Scotty predated Wesley by a couple of hundred years. In the episode "That Which Survives", the Enterprise was beamed a thousand light years away by the ancient Kalandan computer. The ship is about to blow up when Spock instructs Scotty to reverse the polarity of his magnetic probe. Of course, that fixed the problem. I have the feeling that all problems in the Star Trek universe can be fixed by the proper application of reverse polarity.

Yeah, this is great, but....

[Jamey]

Except that it really doesn't help that much!

Hard drives have gotten bigger, and bigger, and *BIGGER* over the last 20-30 years. But they don't *FEEL* that much faster. They've become wonders at streaming huge blobs of contiguous data out - so why do databases need huge steaming bloody chunks of RAM cache? Because the random access times *SUCK* and really haven't gotten that much better!

Capacity has gone from 5MB to 1TB, but spindle speeds have gone from 3600RPM - up to a max of??? 15K RPM for some really expensive drives? Track-to-Track seek hasn't gone up much. Neither has real nor manufacture's claimed throughput rates.

RAM hasn't nearly kept up with CPUs, either, but the disparity is nothing compared to the hold you get when you have to go after some data from the hard drive that isn't in the cache.

It's so bad, I strongly considered putting 3 4GB FLASH modules with IDE adapters (RAID5 - but I didn't study this to see if 2 8GB with RAID1 might be better, or other variations) into my new machine on the PATA header to act as the root drive, holding everything but /home, /var, and /tmp.

Sequential read speed is kinda nice, but I *do* need to do random accesses sometimes! I listen to my nice little 2TB RAID array all the time, as the heads move back and forth singing their little song.

Re:Yeah, this is great, but....

[Anonymous Coward]

Here's the best place I know of for SSDs:
http://dvnation.com/nand-flash-ssd.html

$500 for 32 GB, $400 for 16 - might be worth it to you. Prices are dropping every couple of months, so good to keep an eye on at least.

Bah HD speed

[silas_moeckel]

I can tipple the transfer rate and reduce the average seek time by about the same by using 3 sets of heads. Oh you wanted something thats cost effective please move along. Really though I do not know why they could not use multiple servo motors to at least split the heads already on server class drives, any hardware geeks want to chime in? It seems there is a big push for 2.5 inch SAS drives I cant see why you could not stack some of those platters in a 3.5 and add extra heads and controlling gear? Sure your not speeding up single transfers but your cutting the rotational latency in half and allowing multiple operations at once great for servers.

Re:Bah HD speed

[DaleGlass]

IIRC, the explanation for that I saw was that it's too complicated and does too little good. I think it has even been done, but wasn't successful.

Inside a hard disk it's pretty cramped already. Adding extra voice coils, arm assemblies, etc. is complicated, adds extra heat output, and increases the probability of a failure. A multihead drive would probably cost more than two normal ones and not have much of an edge performance-wise.

Re:Bah HD speed

[Agripa]

I suspect using multiple actuators lost out to multiple drives do to economics. When they were introduced, RAID was becoming much more common in the same markets.

Re:Bah HD speed

[Mal-2]

I have always wondered why drives couldn't be configured with two independent arm assemblies (in opposite corners of the drive so there is no risk of physical collision) that can function simultaneously, which would allow you to either double the read and write speeds, or cut your rotational latency in half, depending on how you used them. Random seek could also be greatly reduced if one set of heads was seeking for the next transfer while the current transfer is still under way. If the electronics are not able to keep up with two sets of heads, they can still be used to reduce latency and seek times. A two-armed 7200 rpm drive should, in theory at least, have latency similar to a 15k drive, and seek times considerably better with intelligent command queueing.

Another method I've wondered about would take far less re-tooling, although it would do nothing for latency or seek times. Read or write (or some mixture of the two perhaps) all of the heads on the drive simultaneously rather than sequentially. This would mean spindle speeds could be kept down but sustained transfer rates could be considerably better than single-platter 15k drives, and the drives would be physically unchanged. Even those 15k drives could be reconfigured to read/write both sides of their single platter simultaneously. The big catch, of course, is the difficulty of keeping ten heads in alignment at the same time. It might be necessary to sacrifice some of this speed to allow the whole stack to be read or written in two or more passes so that heads that slightly missed their mark can have another go at it. Even if it takes three rotations to read both sides of five platters, three is less than ten!

Finally, if each idea is good by itself, why not both together? Two sets of heads to reduce latency and seek times, and all the heads on a given arm active simultaneously to increase transfer speeds -- or the second seek arm could be used to pick up where the heads "missed" on that first pass.

Mal-2

Re:Bah HD speed

[Fweeky]

"I have always wondered why drives couldn't be configured with two independent arm assemblies" They can, it's just not worth it; it's a lot of additional expense and complexity (and thus reduced MTBF) all for a very low volume part, when most people would prefer you to just make a physically smaller, cheaper disk so they can get more of them when needed.

Read-write on all platters at once isn't really feasable because the tracks aren't going to line up reliably; leaving aside imperfect manufacturing, components aren't all going to see uniform levels of thermal expansion or vibration, and even microscopic differences in where each head settles will leave you screwed -- lining up with one track will, most likely, be mutually exclusive to lining up with a second, and get worse from there.

Of course IANAHDM.

Re:Bah HD speed

[Detritus]

Seagate used to make several models of drives with two actuators. They never caught on and were discontinued.

In ancient times, you could get disk drives that had a fixed head for each track, eliminating the need for a head actuator. They were very fast, but their storage density was low.

Modern track densities have made it impractical to have more than one head active at a time. Each active head needs its own independent positioner, servo channel, and read-write electronics.

I've seen mainframe disk drives that had multiple actuators and multiple active heads. That was decades ago, when the VAX was just being introduced and a 70 MB drive was the size of a washing machine.

Solid State Disks

[Anonymous Coward]

How does this fare versus solid state disks? (ie, is their effort a waste of time?)

Hm

[Anonymous Coward]

And from a guy named Stor.

Now all you need is a 3.2km accelerator in your computer.

Re:Hm

[martin_henry]

Now all you need is a 3.2km accelerator in your computer. like an automobile?

Ten Years

[Kuvter]

Working prototype drives should be available within a decade. Sweet, just around the time Starcraft 2 and Duke Nukem Forever come out.

Re:Ten Years

[maxwell demon]

Working prototype drives should be available within a decade. Sweet, just around the time Starcraft 2 and Duke Nukem Forever come out. Yes, but will there be Hurd drivers?

Re:Ten Years

[Kuvter]

Yes, but will there be Hurd drivers? Depends on your definition of hard drive. If you're talking about a hard disk drive, that has moving parts and uses magnetization to store data, then there might not be.

Now that flash drives have started replacing the function of hard drives it'll be interesting to see if we keep calling it a hard drive. If a flash drive can be considered a hard drive, then yes, we'll still have hard drives.

Either way there will be some sort of non-volatile storage device and 640 GB won't be enough for everybody.

Re:Ten Years

[maxwell demon]

Yes, but will there be Hurd drivers? Depends on your definition of hard drive. How does my definition of a hard drive affect the question if there will be Hurd drivers?

Re:Ten Years

[YU Nicks NE Way]

Well, no. But that won't be because the ABI is so crufty and uselessly general -- oh, no -- but rather because the evil software companies have made the software drivers for the unique (and essential) laser retargetters closed source, and the Foolish Stallman Fans will still be busy reverse engineering the retargetters for the pre-prototype versions that were released five years earlier.

Where's my flying car?

[binaryspiral]

Where's my flying car? Damn it - it's still in the labs.

Re:Where's my flying car?

[mjolnir_]

Hey, they promised me a flying car too!

Re:Where's my flying car?

[MobileTatsu-NJG]

"Where's my flying car? Damn it - it's still in the labs."

To be fair, that's not strictly a technical problem. I'm amazed they even let people drive their own personal cars.

Re:Where's my flying car?

[joto]

"Where's my flying car? Damn it - it's still in the labs."

To be fair, that's not strictly a technical problem. I'm amazed they even let people drive their own personal cars.

To be fair, that is a technical problem. There's no reason (beyond technical) for letting people drive their own cars. If you can come up with a technical solution that is better, please do!

Re:Where's my flying car?

[MobileTatsu-NJG]

"If you can come up with a technical solution that is better, please do!"

Technically the FAA wouldn't let me do that. Oh.. hey.. when I put it that way, you're right, it is a 'technical' problem!!

Re:Where's my flying car?

[petermgreen]

flying cars have many problems some strictly techincial, some human factors and some political.

the key techical problem is how to make something that is as narrow as a regular car that is capable of flying, is safe and is not an insane gas guzzler. Removable or folding wings are a possibility (and i beleive flying cars like that do exist) but they mean you need a wide runway to take off just as with a regular light aircraft. They also mean an extra point of failure. Flying without wings is extremely gas guzling.

even if that is solved there is the issue that learning to fly takes much longer than learning to drive and the crowded airspace caused by mass use of flying cars would make things much harder to control. There would also be the political issue of setting up arrangements for people to launch and land thier flying cars near thier homes and workplaces

Re:Where's my flying car?

[joto]

Last time I checked, FAA was standing for Federal Aviation Administration. And I can assure you that they will never make any hindrances for you in creating a car that isn't driven by a person. Now, if you are talking about planes, they might become more interested. But if you can show them that your way is better, I'm sure they'll change their minds. So what was your solution again?

Arguing that we haven't got flying cars because of FAA restrictions on who is allowed to pilot a plane, is about as stupid as arguing that we haven't got teleportation devices because of ethical concerns about what happens at the transmitting end. We haven't got flying cars because nobody has succeded in creating something useful that can be described as a flying car, and we haven't got teleportation devices because it's just too hard for our current understanding of physics.

Perhaps even the idea of flying cars is silly. 100 years ago, people were used to driving horse-driven carriages. And while it's possible to create a contraption that looks like a horse, and make it look like it's pulling a car, we don't do that. Flying is not the same as rolling on a flat surface, and perhaps it's only natural that the two take different forms. There's a reason planes have wings.

Re:Where's my flying car?

[MobileTatsu-NJG]

"And I can assure you that they will never make any hindrances for you in creating a car that isn't driven by a person."

Misunderstanding. I didn't mean an automated car, I meant drivers are so stupid I'm surprised people are generally allowed to drive. Expanding on that idea a bit: Drivers can cause a lot more damage when they're in the air. Poor choice of words on my part, my bad.

"Arguing that we haven't got flying cars because of FAA restrictions on who is allowed to pilot a plane..."

That wasn't exactly my argument, but I'll work with it. Getting a pilot's license is no easy task. It's expensive and it takes many many hours of training. It's nothing like getting a driver's license, and that will certainly prevent Joe Sixpack from owning one. That's not the only problem the FAA would pose, but that's a biggie.

"... is about as stupid as arguing that we haven't got teleportation devices because of ethical concerns about what happens at the transmitting end.

Not really, no. Not even in the same ballpark, really. If teleportation carried a risk of randomly causing explosions around the planet, well then we'd sort of be in sync with this silly comparison.

Re:Where's my flying car?

[joto]

Getting a pilot's license is no easy task. It's expensive and it takes many many hours of training.

In Norway, you have to be 18 to get a drivers license. You have to be 17 get a sailplane pilot license. At 16, you can get a license for a hangglider or a paraglider. Very few people can afford their drivers license, unless their parents pay for it. All of these pilot licenses are much cheaper than getting a car drivers license.

In order to get a license for an ultralight motorized plane (microplane), you have to be 18, but it's still about half the price of a car drivers license. The cost of a pilot license for a small motorplane (i.e. Cessna, Piper) is about 1 1/2 - 2 times the cost of a car driving license.

Getting a car drivers license is neither particularly easy nor cheap. But most people need a car drivers license, which means that they will put up with the studying and the economic cost. Pilot licenses are viewed as a luxury thing, and are therefore perceived as more difficult and more expensive, even though they're not.

I'm not sure how these numbers compare to US numbers, I would expect getting a drivers license is at least a bit cheaper there, but I believe the numbers are still somewhat representative. Of course, most pilot licenses mentioned would be more expensive if the people involved in training viewed it more as a job than a hobby, but the numbers would probably still be correct within a factor of two.

It's nothing like getting a driver's license, and that will certainly prevent Joe Sixpack from owning one.

If Joe Sixpack can afford his brand new SUV, he can afford to get a pilot license. The problem is that pilot licenses are not useful to the average person, it's a hobby, or for some very few: a profession. A flying car would change that, but we haven't got flying cars, and that's the main problem, not the FAA.

Once you can come up with a low-noise, cost-effective, small-footprint, safe, VTOL plane with useful payload, I can assure you that the FAA would respond to that. But it doesn't exist, and there's no reason for FAA to change regulation in order to make it easier for people to imagine a future with flying cars.

Re:Where's my flying car?

[MobileTatsu-NJG]

"In order to get a license for an ultralight motorized plane (microplane), you have to be 18, but it's still about half the price of a car drivers license. The cost of a pilot license for a small motorplane (i.e. Cessna, Piper) is about 1 1/2 - 2 times the cost of a car driving license."

Umm.. okay. In the United States, the expensive part of getting a pilot's license isn't getting the license. It's the many many hours of training, both book studies and actual flight experience. We're talking in the neighborhood of $5,000 and a minimum of 40 hours of flight time to get that license. Getting a driver's license is NOTHING like that.

"Getting a car drivers license is neither particularly easy nor cheap. But most people need a car drivers license, which means that they will put up with the studying and the economic cost."

I spent roughly $400 on my driver's license, and that included the price of the car, the gas, and the fee. (For the record: It was a VERY cheap car. Hehe.) The $5,000 figure I mentioned earlier came from some information I obtained when I worked at a company that built airplanes. Since I was staff, they were willing to let me earn my flight time more-or-less at-cost, which was in the neighborhood of $1,500 to $2,000. Now, I can happily accept that in other countries, it may be considerably cheaper and more practical to get a pilot's license. But here in the USA, it's very expensive, and that doesn't include the cost of buying or renting a plane to use this license with.

"If Joe Sixpack can afford his brand new SUV, he can afford to get a pilot license."

No, this really isn't true. He may be able to round up 5k, but that's not the same as 'affording' it. Plane rentals and fuel are expensive.

"A flying car would change that, but we haven't got flying cars, and that's the main problem, not the FAA."

The technology to build a flying car is already there. I should know, I helped design one. It's actually possible to build a small plane that'll fold up into roughly the form factor of a car and 'drive'. And that was just our simple model, there's a few prototypes out there of stuff that's far more interesting and borderline scifi-ish to look at. The problem with these things, though, is that they're still well over 100k, difficult to get certified for flight AND for driving, and there aren't enough people with pilot's licenses ready to plunk down the cash. They cannot be mass marketed because the pilot's license requirements are so steep. Even then, there's concerns about that many flying vehicles in the air. (That one actually is a technical problem, better auto-pilots etc.)

We've been able to build flying cars since the 50's. Technology isn't the problem.

Re:Where's my flying car?

[joto]

We're talking in the neighborhood of $5,000 and a minimum of 40 hours of flight time to get that license. Getting a driver's license is NOTHING like that.

NAK (Norwegian pilot association) gives the cost of a typical pilot license for a small motorplane at 45 hours flight time, total cost NOK 55655, or USD 9566. The local driving school advertises an intensive two week driving course with theory, exams, all required courses, a total of of 30 hours driving time (most of it is required by law), and all state fees, at a grand total of NOK 26865, or USD 4617, assuming you pass (most people don't). It seems a Norwegian driving license is EXACTLY like that.

It's actually possible to build a small plane that'll fold up into roughly the form factor of a car and 'drive'.

While I agree that such a contraption can be called a flying car, it's not what most people imagine when they want a flying car. They want something that is as useful as a car, that can fly. What you described is something that is totally useless, but can still (technically) be described as a "flying car". We've seen these cars with wings (or planes with car chassis), and we're not impressed.

We've been able to build flying cars since the 50's. Technology isn't the problem.

Oh please. If that was true, Michael Jackson would own one in Neverland, and ride around in it all the time (at least until he was arrested). Rich people are already lining up to get a flight on a commercial space plane. Where are their flying cars? Don't tell me they aren't interested, because everybody else is. And why doesn't other organizations with money use them? Such as the military, rescue service, etc?

Re:Where's my flying car?

[MobileTatsu-NJG]

"It seems a Norwegian driving license is EXACTLY like that."

Fair enough. But here in the US, you pass a written and a driving test and you're good to go. If there is a fee, it was so negligible that I don't even remember what it was. As I recall, nobody I knew had trouble getting a license. The hard part was getting the car. The license came a week or two later. Maaaaybe a month in some cases. I actually had a tougher time getting my Ham Radio license than I did my driver's license.

For the record, I feel silly bragging about this. I like the Norwegian system better. Anybody can get a driver's license here... and it shows.

"While I agree that such a contraption can be called a flying car, it's not what most people imagine when they want a flying car. They want something that is as useful as a car, that can fly. What you described is something that is totally useless, but can still (technically) be described as a "flying car". We've seen these cars with wings (or planes with car chassis), and we're not impressed."

Yes, I know, we'd all like to be driving the Delorean from Back to the Future. Funny thing is, a car kind of like that exists in prototype form, today. It has four turbine engines that extend out from the body of the car. It has an aerodynamic form that has a hint of wings, but still fits within the size of a large car or SUV. It could even hover. Problem is, you still need a pilot's license, and it still hasn't been cleared to be sold. And no, you don't need to step out and affix wings to it. But, that's a different company working on that. What we did was more like you describe, only it did act as a car and the wings folded out on their own. It was meant to be used for making trips in between cities without needing to rent a car in the mean time. You wouldn't go drag racing in it, and it was a little... Demolition Man looking, but it would have sufficed so long as your interests didn't include landing on buildings. Heh.

"Oh please. If that was true, Michael Jackson would own one in Neverland, and ride around in it all the time (at least until he was arrested)."

If memory serves, he has a helicopter. Anyway, that rationale's amusing to me. Right now you can hire pilots to take you where you want to go. I'm not talking private jets, here (although that's what rich people often lean towards). I'm talking about going to a local airport and hiring a guy with his own prop-job. Pretty reasonable, really. Some companies actually use these services to quickly ship things like expensive computer systems from one city to another. Still, though, the vehicles I'm describing are not very fast compared to say a commercial jet. At best you'd buy yourself some privacy and more convenient schedule.

"Where are their flying cars? Don't tell me they aren't interested, because everybody else is. And why doesn't other organizations with money use them? Such as the military, rescue service, etc?"

Helicopters are more practical. They don't require a runway, etc. We're also not talking about vehicles here with incredible speed capabilities. As I recall the specs I heard about were somewhere around 200 mph. It miiiight have been 300, but I wouldn't swear to it. The engine that would fit in one of these vehicles wasn't not very high power and its hauling capacity was limited compared to a full on plane. At best it was a 4-person transport. Donald Trump would laugh at that.

Tough to market... I suppose in that respect you could call it a technical limitation. The reality, though, is that it gets a lot more attractive when you don't have to spend tons of money and time to get a license. (Not that I'm faulting the FAA for that...) Or, perhaps, if it had an automated piloting system (and that system scaled up to thousands or millions of users across the country) the FAA would be happier with it. Meh, I dunno. I guess the point is you're not entirely wrong about the technology behind it, but the big fear right now is these things falling from the sky. Nobody wants to be liable for that.

a decade?

[mrobinso]

That's a decade before working prototypes are available.

Then there's the question of when they'll be available to consumers at a price that doesn't require selling one's children in order to afford one. Tack on another decade. Considering the size of flash drives are doubling every couple of months, in 20 years I should be able to carry around a couple of terabytes on my keychain. And I want spinning metal disks made of rare-earth material why?

Gadolinium is worth about $170US a kilo right now. Soon as the free markets find out it can be used for something other than in a nuclear submarine that'll go up. I remember when diesel was 1/4 the price of gas until diesel engines became more popular. How'd that work out for us? Oh. In passing, gadolinium exposure has been associated with a higher-than-normal incidence of kidney failure.

Um... I'll pass.

its the physical limitations, stupid

[Gothmolly]

So the drive still runs at 10k rpm, and still has metal platters that heat up, and still only has a 16mb cache, but you can read it 100x faster..... resulting in exactly nothing.

It's time for a new form factor, think a stack of quarters rather than a pile of LPs. I want a 1/2" diameter drive, with 16 platters. Then you can talk about media speed.

how noisy is it?

[Karma Sucks]

How noisy is it? For me it's far more important that a hard drive be quiet and well-behaved than it be fast. From what I remember from James Bond and other movies, lasers are pretty damn loud.

Re:how noisy is it?

[Anonymous Coward]

Lasers don't make any noise. Maybe the huge scifi lasers in James Bond do, but not ones in real life.
Have you listened to a laser pointer? The only noise from the drive would be the disks spinning and the head moving.

Re:how noisy is it?

[SacredByte]

From what I remember from James Bond and other movies, lasers are pretty damn loud. Espially when they're in outer space.....

arm motion?

[commodoresloat]

I'm willing to stop moving my arm for 40 femtoseconds if it will help.

Gadolinium - It's not just for harddrives anymore

[weinrich]

the secret of the current work's success lies in its disk's materials -- gadolinium, iron, and cobalt.

Interestingly, this same material is being used in a variety of other fields, including medical; specifically in Magnetic Resonance Imaging (MRI).

From the article: Engineered Nanomaterials Improve Magnetic Resonance

Two groups of investigators in Europe have developed engineered nanoscale materials that enhance images obtained using magnetic resonance (MR) imaging. With further development, these nanomaterials have the potential to improve the detection of early stage cancer.

Rotational latency

[danwat1234]

Yea, how is this technology supposed to help the hard drive industry? There is still the latency with waiting for the desired section of a track to come around, not to mention the Servo seek time delays! How does it benefit? Read/write heads can already read and write data off a track as fast as the platter can spin!

Where does it say that they were able to READ?

[laing]

I have RTFA and cannot see where this group has successfully read data from the drive with a laser. All it talks about is writing.

Re:Where does it say that they were able to READ?

[xmartinx]

yeah, i also wonder how that's supposed to work. i can't think of a way to do that using a laser and doing it fast.

Re:Where does it say that they were able to READ?

[CryptoDavid]

Presumably using the Magento-Optic Kerr effect; light reflected from the surface of a magnetic material is circularly polarised. It sounds like they're recording using this effect in reverse, so read and write speed should be comparable.

However, TFA says 'managed to transfer data at intervals of about 40 femtoseconds' which is surely very misleading - it implies they wrote data at 25GHz. Given their spot size of 5um, that would require the medium to be moving at 125,000 m/s.

Of course, the actual paper says something quite different; their key achievement is being able to change the magnetisation of a domain in just 40 femtoseconds.

Re:Where does it say that they were able to READ?

[whit3]

Yes, the original article omitted READ operation. The old NeXT magnetooptic drives
used the polarization-rotation trick, and never got very fast. And there's nothing
in the article to indicate the READ operation is up to speed.

Not a problem, though; just scream the disk at 30kRPM for write, and slow down
to 3 RPM for read. You'll have the comfort of hearing your hard disk bray and
whine like a jackass...

Re:Where does it say that they were able to READ?

[xmartinx]

wehaa, back to hard-disk-races :D

Fast!

[CyberPhoenix]

I am sure the Vista users are wishing they had a 100x faster hard drive.

A variation on a bad joke

[Anonymous Coward]

3D Realms is reportedly adding special features to Duke Nukem Forever that will take advantage of these advanced hard drives.

The business process has failed

[billcopc]

Working prototype drives should be available within a decade

Just in time for the bargain bin at Best Buy.

Seriously, in the world of computing, if you have a working prototype, you need to haul serious ass and get the product out ASAP because things move so quickly. In a decade, we'll have bigger and faster magnetic drives. They used to say we'd run into the density barrier, then someone tried perpendicular recording and breathed a little more life into the technology in a timely manner.

The other issue is: what kind of bus can handle those speeds ? If they're not shitting us with their 100x claim, that makes this hard drive faster than most RAM in use today. What the hell ? This is either completely revolutionary or a crock of shit. Call me cynical, but I'm leaning toward #2.

Re:The business process has failed

[Anonymous Coward]

You need to do a little more research on your facts before you post your shoot from the hip nonsense. I agree with your stance on 10 years.

FYI: SATA can handle approximately 20X the buffered read speed of just one disk. Bus speeds are *much* higher than disks can output right *now*. Hell, EIDE can handle about 10X for a 7200RPM disk. You don't seem to have a clue how *much* faster RAM really is. Sorry man, but 100X ain't gonna get you close to RAM speeds.

Ideal for games

[NightFears]

Working prototype drives should be available within a decade.

Surely it it will have enough space to hold the entire Duke Nukem Forever... if you store it on a WinFS partition.

I heard...

[settrans]

...that they'll be selling these on the next Pink Floyd tour with Duke Nuke Forever preinstalled its WinFS-formatted partition. Also on sale will be pocket cold fusion generators, cell phones running Skype, and CmdrTaco tshirts.

A decade??

[Anonymous Coward]

The speed is awesome.... But latency of one decade is crap and i will not stand for it.

Article

[Derosian]

Is no one reading the article, or even the headline. The best thing about this is faster read and write. The one biggest slowdown on modern computers. Disc compression is great and all, most space more fun, but when I can load Windows Vista in under a minute that will make me very happy.

Re:Article

[Anonymous Coward]

Or load up your database of stale Microsoft jokes in under a day.

Damn!!!

[donnacha]

... I knew this would happen if I went ahead and ordered my Macbook Pro!

Faster?

[Gorimek]

According to the article, it will take several years to write using this technology, even if you start right now. I'll stick with my current HD, thank you very much.

Then all we need is a tracking system in space!

[Wain13001]

And we can vaporize a human target!

Different polarizations

[GanjaManja]

Laser light *may* be polarized in some certain direction, if the laser is designed to do that. But more importantly, there is circularly polarized (electric field oscillation direction rotates with time) and linearly (it does not rotate with time) polarized light. Elliptical pol. is somewhere between these two extremes.

Circular pol. corresponds to a spin-flip (1/2 x integer spin change) when absorbed by a bound electron, and linearly pol. corresponds to an integer spin change (if i remember correctly). I would guess this is how they managed to correlate light's polarization to magnetic orientation. Or perhaps they simply use counter-clockwise circularly pol. to flip the magnetic bit one way, and clockwise circular to flip it the other way.

short answer: yes, there are 'degrees' of the polarization of light.

Re:Different polarizations

[NeilTheStupidHead]

Thank you to both that replied. I have never really dealt extensively with any technology that uses lasers so I don't know that much about the mechanics... well I do work with fibre now and again but laying cable doesn't require knowledge of how a laser works to beyond the basics.

in the meantime

[Khyber]

We're already producing solid state drives with practically nil seek time, no moving parts, and quite potentially a much longer lifetime than a moving-parts drive.

But most of the /dev/null emails are spam...

[patio11]

Poor Scientologists, they're about to receive billions of letters from profit-seeking quackpots pushing quasi-scientific remedies with no chance of working to try to hook the gullible and stupid...

Wait, there is some poetic justice here.

Size Obsession In Reverse...

[FlyingGuy]

Back when I started out in this biz, the 5.25" full height 10Meg hard drive was all there was, and it was good, and I forget how many platters that damn thing contained.

Then came half height 5.25" drives ( the Seagate 20meg comes to mind ) and now we have launched into terabyte territory on 3.5" single platter drives

When the multi-platter drives got bigger and bigger they started writing disk drivers with "elevator" algorithms so that data was written to all the platters simultaneously and therefor you could read with ALL the heads at once and get a very large chunk of data at once.

So for small devices like MP3 players, laptops and the like we need ever smaller drives with ever higher capacity, but for servers and large storage devices why not go back to larger form factors. You could loosen up the track density a little and make them more forgiving, yet increase the capacity by say 20x. 5 double sided, 5.25 inch platters all spinning on the same spindle. I don't think the motors would have to be that much more powerful since in a storage farm you don't really care if it takes 5 or 25 seconds for the drive to get up to speed, what you care about is storage capacity and longevity.

One of the things I don't know and would love to have an answer to is how many times around does a head have to see the same bits to decode them correctly.

What a waste of research cash and time

[diorcc]

With SSD drives reaching the consumer market, why bother with a technology thats slower than them before its even completed?
100x faster is nothing... look for new ways to store data, living organisms, whatever--- why must we still use the spinning disc? ;)

Yeah at the moment SSD storage is expensive but I expect that to go down with consumer demand going up.

?- Swapping != Thrashing

[PingPongBoy]

More RAM? Heh. If the claims of 100x read write is true, swap = RAM. After all, with light, all you have to do is shine at multiple points to eliminate seek. Heck!

Big drives

[Anonymous Coward]

Don't overlook business needs.

We needed locally attached storage for one machine, we bought an Addonics card and box and some 320G SATA drives, totaling about $900 for 1280G. We did that because with the 1TB drives we could expand to 12TB at need. We need the ability to store about 50TB of video but we don't need speed so we bought a big box of USB2 external drives, it should come to about $16,000, of course we are buying in blocks of 5TB so final cost might be lower. Since we just plug in a different one when we need more space, the expandability is practically infinite. We needed centralized storage for critical data that we could use at a block or file level and we got 6TB for about $10,000. We can double that anytime we like, maybe quadruple without problems.

We needed and expect to continue to need more and more storage. We handle financial data so we keep practically everything (the video is surveillance) and expect to continue to need more and more indefinitely. We do need some pretty powerful processors and for the most part we satisfy our need for processing by buying more boxes and dividing tasks when the need arises. (Virtual Machines are our future for this.)

Honestly I don't think that our processing power needs are going to more than double in the next ten years, but I expect our storage needs to continue to expand dramatically, maybe doubling every two years. Do we need faster storage? No, not really, we put in SCSI where we need speed, and with RAID 5 and 4 or 5 drives we haven't seen any problems yet. At my last job though, they did need more speed and they couldn't get it, not with fiber and RAID0 SCSI. I think it was due to poor planing and the limitations of SCO (yep, SCO) but that wasn't my area. I can tell you though, that if they could have purchased 1TB with massive improvements in speed, you could have named anything with six figures as your price and they would have thought they were getting a bargain. This _will_ matter to businesses.

* The DAS and centralized storage figures are before RAID5, approx 900G and 4.5TB after.

Re:Big drives

[Anonymous Coward]

we bought a big box of USB2 external drives

I can pretty much predict that you'll be pulling your hair out within 2 years. I'm betting that you'll find failure rates on external USB enclosures to be around 5-10% per year (maybe less - but still enough to drive you up a wall). Either the power supply will fail, or the power connector, or else the enclosure will simply cook the hard drive. (I've seen very few enclosures where the HD could be used for more then short spurts without heat issues.)

We needed locally attached storage for one machine, we bought an Addonics card and box and some 320G SATA drives, totaling about $900 for 1280G.

And if the data is irreplaceable - never use RAID0. Especially if you don't have the data backed up onto a another piece of physical media. For local scratch space, RAID0 is fine. But you'd better be doing some sort of backups if that working data is important.

We handle financial data so we keep practically everything (the video is surveillance) and expect to continue to need more and more indefinitely.

That statement, along with what looks like trying to go the low-cost el-cheapo route... makes me hope you're not working with any financial institution that I deal with. Seriously, invest in some SATA SAN units (which aren't that expensive on a per-GB basis) from a commercial vendor. They're very good for high-capacity needs where you need a lot of storage but not top of the line speed.

Re:Big drives

[Anonymous Coward]

Humm.. I seem to be having to repeat myself. Maybe re-read my original post so you don't look silly eh?

I don't care if the USB drives go belly up. They're disposable data. We don't need the data on them except very occasionally and when we do it is almost always when they are still in service. If one of them is needed after it is pulled out for rotation that once a year it might happen, and that one time just happens to be one of them that failed then we send the drive out to someone to be repaired. We've got like 60 days before it is important and it is never critical. No, I don't expect to tear my hair out, I'll just throw the drive away.

You missed the note at the bottom, we're using RAID5 for all important data. Actually we do so many backups that we could probably get away with RAID0 but we don't even talk about it, see there was this one time (before I joined the company) that somebody didn't understand what... well now we have experience and don't do RAID 0. I do describe the space in raw capacity because we optionally could use RAID1 or RAID10 or RAID50 and all those change the writable space. If you were my boss you'd get two sets of numbers, total and then I'd focus in on "enterprise class" space which includes figures for writable space after a redundant RAID (redundant Redundant.. and this is redundant too now.)

Funny you should mention a SATA SAN unit, thats what the 6TB unit we got for $10,000 actually is. If you're in the business then you'll be shocked but they're a new company with great hardware and they don't normally sell SANs. We were planning on using EqualLogic or LeftHand before we tested their systems, but it comes out to about a third the cost for at least as good on the hardware side and the best support and warranties we could ask for.

New Tapes

[pseudosero]

Digital Tapes. Gonna have to re-record the world again.

I've been hearing this stuff for 15 years

[AbRASiON]

There's always always a new disk technology around the corner, something finally bigger, faster, better than magnetic, it's always 5 years away.

Honest guys! the 2TB holographic drive, with 3 lasers (top,front side) into something the size of a sugar cube is coming!
(I heard that one nearly 10 years ago)

Seagate, WD, Fujitsu, Hitachi would / will / have / may band together to hold any tech like this back or at least 'stagger' the release of it.

If they ever _EVER_ release a disk under 500$ US which is say 4x bigger than the current largest (say a 4TB now) and obscenely fast it will sell like absoloute hotcakes, 500$ or not - seriously.
It will decimate their current inventory and manufacturing tools overnight, I for one would have absoloutely no hesitation in dropping that kind of cash - maybe even a grand.

So, to summarize..... I guess one day this might come - but by the time it does, if it does you can be darn sure they'll make sure it's only marginally better than what they just released earlier.......

Yes, reading is easy

[mbessey]

Reading is no problem - standard magneto-optical drives can read the magnetization of a disk with a laser very effectively. As mentioned above, the real breakthrough here is being able to write to the disk quickly. Current M-O drives take 2-3 times longer to write data than to read it.

It's been tried

[rrohbeck]

There was dual actuator drive on the market a number of years ago. It turned out to be so much of a niche product that it cost more than two competing standard drives, which give you the same throughput and twice the capacity.

Modders stay away

[HTH NE1]

Gary lets us know about research out of the Netherlands that has succeeded in reading and writing a hard disk using polarized laser light.

To those modders who like installing plexiglass/perspex windows on their hard drives to see inside them while they are operating, might I suggest they keep their Dremels away from these models?

Article deprecates disk drive performance

[Anonymous Coward]

StorageMojo points out that last generation 15k drives already write over twice as fast as the article claims they do.

Add to that the points that in 10 years areal density will be up at least 100x, that write performance is only a part of total disk performance and that you'd need 75k drives to take advantage of the increase.

One can only conclude that the slow summer silly season is upon us.

Hard drives in 10 years?

[jshriverWVU]

Not to sound like a troll. But do you think we'll still be using hard drives in 10 years? I was hoping by then we'd have something like solid state drives at an affordable rate to be more realistic. But to fantasize a bit, maybe HD's will be replaced by odd shaped crystals and some 3d rotating laser. Even that isn't too far fetched. There was an article a year or two ago, where NASA had a research project on the ISS where they were building custom crystals (something about zero G and crystalization) anyway, the crystals had many faces that the researchers where going to use to burn data on each face. Wish I knew where I read that article, it fascinated me at the time.

OT: About your sig ...

[Anonymous Coward]

What are you doing every time you visit a web page? WWW = double u double u double u = 2*(2+1) 2*(2+1) 2*(2 + 1)

1. How (or why) do you equate u with 2+1?
2. Why do you have spaces around the last "+", but not around the first two?