Sign in with
Sign up | Sign in
Your question
Closed

Laser Heat Used to Make HDD Write Transfers Faster

Tags:
  • Heat
  • Hard Drives
Last response: in News comments
Share
February 9, 2012 9:54:57 AM

Nice, bring it on!
Score
10
February 9, 2012 9:57:33 AM

good, now where's my 1PB RAID 1 array? :D 
Score
7
February 9, 2012 9:59:32 AM

Wait... what about read speeds?
Score
20
February 9, 2012 10:05:40 AM

isn't that how minidisc works/worked?
Score
-2
February 9, 2012 10:09:15 AM

dcdc1isn't that how minidisc works/worked?

ah - didn't read it properly - seems a bit like minidisc but without the magnetic head.
Score
-4
February 9, 2012 10:27:58 AM

800 degrees Celsius O_O
If a write a full disk in one go then I can make a BBQ as well, great news!!! xD
I wonder what would happen if a make a copy of a full RAID, ROFL
Score
7
February 9, 2012 10:48:25 AM

will the new drives be resistant to floods and price increases in contrast to all tech trends?
Score
0
February 9, 2012 11:05:12 AM

imagine how HOT a hard drive would become using this tech... at least there would be less of a chance of a head crash i imagine now how would it go about helping reads?
Score
0
February 9, 2012 11:13:05 AM

This sounds promising.

I wonder is hard drives have been traditionally rated by rotational speed (5400/5900/7200/10000) RPM's
then, how would they be rated now when/if this technology makes it to the consumer?

Will this laser allow for multiple recordings on hard drives?

What about Disk Defragmenting?

Like some asked: What about read speeds?
Would it boot faster than a hard drive?
Would speed be comparable or better than SSD's?

More importantly:
When this drives make it to the consumer, Will I be able to keep all my body parts?
and... Will my wallet suffer?
Score
5
February 9, 2012 11:24:30 AM

This is nice, but there is a problem: they found out how to write data, but they have no applicable way yet to READ it. They are using a scanning tunneling electron microscope for that and these are room sized devices. So don't get too excited, it will take a long time until we see such devices on shelves.


Score
3
February 9, 2012 11:27:39 AM

Very interesting. But do I get this right, it will "only" speed up the writing of data, not the reading of course !?
Not that faster write speeds are not appreciated !!
Score
2
February 9, 2012 11:44:41 AM

@freggo

There is no read yet, this technology is still in diapers, basically they were just experimenting with lasers and "reading" the results with ST microscope. It's more just a demonstration, it is way too early to talk about practical application. There are many more problems to be solved.
Score
4
February 9, 2012 11:47:21 AM

Cool.
Score
2
February 9, 2012 11:55:21 AM

While this technology would be great for people like me who need sequential throughput, I fail to see how this would help the average person because it will still have the same seek time as other platter-based drives today. Remember, sequential throughput on a HDD is not all that bad (120-150MB/s); it is when things are not sequential that thing go to hell (30-60MB/s).

Also, I find it hard to believe that it would take less power to heat a platter than to aim a small radio at it. Besides, would this not wear out the media having so many localized temperature variations? I mean, it is cool tech, but hard to believe it would work 'as advertised'.
Score
5
February 9, 2012 12:15:13 PM

Very interesting, all we need now is for the rest of the system to work at that speed, that fast your RAM and CPU would be a bottleneck
Score
0
February 9, 2012 12:29:50 PM

dcdc1ah - didn't read it properly - seems a bit like minidisc but without the magnetic head.

Not going to lie, first thing came to mind as well.

This is laser only for writing, while MD used a laser for heat and magnetic to write then read with the laser alone.

Still make me think of MD for sure.
Score
1
February 9, 2012 12:44:27 PM

How does this technology relate to HAMR (Heat Assisted Magnetic Recording)?
Score
2
February 9, 2012 12:45:15 PM

caedenvRemember, sequential throughput on a HDD is not all that bad (120-150MB/s)

Only the fastest mechanical drives in the world are that fast, and only when the data is both sequential and at the edge of the platter for maximum rotational velocity on the platter. This is obviously a fairly rare occurrence.

Quote:
it is when things are not sequential that thing go to hell (30-60MB/s).Also, I find it hard to believe that it would take less power to heat a platter than to aim a small radio at it.

Generating magnetic fields actually requires considerably more power than a picosecond pulse of a laser.

Quote:
Besides, would this not wear out the media having so many localized temperature variations? I mean, it is cool tech, but hard to believe it would work 'as advertised'.

The amount of energy being delivered is very small. Metal is remarkably elastic, and not all metals are subject to fatigue (getting weaker after repeated small stresses). This will probably be one of the easier problems for them to solve.
Score
3
February 9, 2012 1:11:30 PM

wonder how it feels to take an assumed scientific fact that has been believed to be true for over a hundred years and make it do the opposite of what everyone has been saying it would do. Granted, it took a lot of researchers to do it and I don't think that they turned the principle on its head, but they still got it to do the opposite of what was thought. pretty cool.
Score
2
February 9, 2012 1:18:57 PM

Interesting. Sounds like this has the most promise to increase write speed, but it sounds like the net effect is the same as if data were recorded magnetically. Therefore, it will likely still need to be read magnetically - which I would assume for the time being means that read speeds will not increase.

Also, this article says nothing about commercialization of the technology, and my bet is that it is years off. I hate to say this, but I am not getting my hope up.
Score
2
February 9, 2012 1:35:11 PM

This is actually somewhat old news. I remember writing a paper on this for one of my graduate courses back in Spring of 2010. The cool thing is that, in addition to the increase data transfer rate, the data density is also increased by nearly a factor of 10! If we can implement this technology in a practical fashion, we would see 100TB drives appear on the market practically overnight.
Score
3
February 9, 2012 1:47:08 PM

Well, for me the low power, low heat is more important in HD than pure speed. Speed is nice, but not if there is big heat penalty.
If I need more speed I will chose SSD...
Ofcource if the heat is no issue, this can be guite interesting?
Score
2
February 9, 2012 2:27:05 PM

jellicoThis is actually somewhat old news. I remember writing a paper on this for one of my graduate courses back in Spring of 2010. The cool thing is that, in addition to the increase data transfer rate, the data density is also increased by nearly a factor of 10! If we can implement this technology in a practical fashion, we would see 100TB drives appear on the market practically overnight.


OK, I'm going to bed now, wake me up as soon as this drives are on sale....
Score
2
February 9, 2012 3:30:11 PM

When will these hard drives become available?
Score
-1
February 9, 2012 3:39:01 PM

Am I the only one wondering if this could allow them to decouple the write and read processes from each other? Having the laser write array opposite the magnetic read arm would allow you to read and write from the drive at the same time wouldn't it? Maybe add some slight delay to allow the onboard circuitry to register where the writes were, but the head wouldn't be competing for either reading or writing.
Score
3
February 9, 2012 4:20:40 PM

freggoVery interesting. But do I get this right, it will "only" speed up the writing of data, not the reading of course !?Not that faster write speeds are not appreciated !!

At this point there's no way to read a magnetic field using a laser. So yes, only write speeds could be improved. I don't think write speeds are an issue at this point anyways...it's *random* writes that are real PITAs because HDDs are tracked mediums.
Score
2
February 9, 2012 4:21:54 PM

tuluiAm I the only one wondering if this could allow them to decouple the write and read processes from each other? Having the laser write array opposite the magnetic read arm would allow you to read and write from the drive at the same time wouldn't it? Maybe add some slight delay to allow the onboard circuitry to register where the writes were, but the head wouldn't be competing for either reading or writing.

I think it was proposed a long time ago that dvd's could be burned faster if there were multiple heads doing the writing. Don't remember what was wrong with it, but clearly there was a problem otherwise we'd see it today.
Score
-2
February 9, 2012 5:28:51 PM

So these drives are going to have write speeds of 1000mb/s but only 100mb/s read speed, well if the price is right it'll be worth it
Score
2
February 9, 2012 5:47:24 PM

Meh....it would be good for hybrid drives. No write wear on the SSD portion with ultra fast HDD writing, while still leveraging the SSD's lightening quick read capabilities. In the end, this is nothing more than a new technology that will prolong the life of HDD storage.
Score
-3
February 9, 2012 6:56:53 PM

Since hard drives are mechanical devices, there would still be seek times and rotational latency to consider in determining throughput. If there is hope for this technology using conventional hard disk design, it would lie in its ability to write more bits per square inch than magnetic write heads can do.

If you can write a million bits on a single track instead of 100,000 using today's magnetic write heads (and assuming the magnetic read head can still read the bits at that density) then you could see a 10x improvement in throughput for reading or writing just that track. You would also see benefits from fewer seeks since more of your data is likely to be concentrated within fewer tracks.

So if you can write 5-10 bits or more in the same space as you can write a single bit today, then I am very impressed with the potential of this technology. If not, I'm not sure I see the real benefit.
Score
2
February 9, 2012 8:11:15 PM

Nice to read about more possible solutions to how slow storage is. However, we have been hearing about these better storage technologies for decades, but only flash based SSDs have really come out of it for consumer storage.

We have small chips based on some of the super storage tech like MRAM, but we still don't see any devices utilizing this stuff. What is keeping them? I don't care what problems there are, there is no way that every single technology that could beat current storage isn't working yet do technical problems. Is something or someone(s) stopping them?
Score
0
February 9, 2012 10:36:58 PM

caedenvWhile this technology would be great for people like me who need sequential throughput, I fail to see how this would help the average person because it will still have the same seek time as other platter-based drives today. Remember, sequential throughput on a HDD is not all that bad (120-150MB/s); it is when things are not sequential that thing go to hell (30-60MB/s).Also, I find it hard to believe that it would take less power to heat a platter than to aim a small radio at it. Besides, would this not wear out the media having so many localized temperature variations? I mean, it is cool tech, but hard to believe it would work 'as advertised'.


this tech applied to todays drive capacities allows for 10-20 times the current size. so you would be getting 40-80tb, as most people have no need for all that space, its conseveable that most people would never realisticly fill it. so the write speed would be a great improvement. and if read speed scaled with the increased size (not as much as write) we would have 1gb-2gb read speed (based on my current 1.5tb read speeds).

now if windows was installed in a way that would put a 100gb chunk in the beginning just for os and load a sequential read, the seek time would be more or less un needed, possibly makeing them faster/cheaper also than an ssd for that task...

hell if you figure people will never use all that hdd space, you could force programs to all write sequential, instead of now, where they know they cant write sequential all the time. the innital 15-30 what is it, ms seek time would be unimportant, considering you will have the choice of a 256gb ssd or a 40-80tb hdd that preform at about the same speed, but the hdd also allows mass storage. it would make a ssd a very hard sell.

Score
1
February 10, 2012 12:38:03 AM

Can I get a cup of coffee with these heat lasers?
Score
-1
February 10, 2012 2:36:42 AM

It heats an extremely tiny space up for a very short time. The drive shouldn't run very hot, maybe not even as hot as current drives do.
Score
0
February 10, 2012 4:50:22 AM

Does that means we dont have the driver HDD read hand? If thats the case we could have a HDD that is much more prone to shock resistance.!
Score
0
February 10, 2012 8:42:00 AM

THis is not new technology. My LS-120 used laser technology to record on floppy disk. It had 120MB capacity, and fast write speeds. also I could read standard floppy disks really fast and format those at 20MB+ if I wish. THere was an LS-240 version that was to come out. Point is laster to read/write magnetic bits has been out since 1996/7.
Score
1
February 10, 2012 8:44:41 AM

the LS-120 floppy was 20% faster then when it was mechanical..... so I guess hdds would have great speed increases, also since they rotate faster. And maybe rotation speed can be increased making even more speed improvements.
Score
1
February 11, 2012 11:15:21 PM


shqtthTHis is not new technology. My LS-120 used laser technology to record on floppy disk. It had 120MB capacity, and fast write speeds. also I could read standard floppy disks really fast and format those at 20MB+ if I wish. THere was an LS-240 version that was to come out. Point is laster to read/write magnetic bits has been out since 1996/7.


LS-120 has absolutely nothing to do with this tehnology. LS-120 only used laser to GUIDE a magnetic head.
Score
1
Anonymous
February 28, 2012 7:11:02 PM

Just one thing. Laser emission diodes becomes weak during the time. It means - Such HDDs will have much shorter lifetime and the risk to loose data will be much higher than for ordinary HDDs.
Score
-1
February 28, 2012 8:06:40 PM

scepticistJust one thing. Laser emission diodes becomes weak during the time. It means - Such HDDs will have much shorter lifetime and the risk to loose data will be much higher than for ordinary HDDs.


Yes, that's why my DVD writer from 2004 is working without any problems right now, as is my original playstation and some CD players from the late 90s that still see common use as they have for around two decades.

Maybe some laser diodes get weak, but it's obvious that not all do. Also, I have an old laser pointer that sees weekly use and looks no weaker than when I was first given it several years ago.

Hard drives only need to last up to five years or so for most people. I would prefer that my drives last ten years or more, but that isn't necessary. If these hard drives can't last ten years, well I guarantee that they wouldn't be sold in high volume unless they had at least a 3 year MTBF. Most people that hold on to systems for long periods of time aren't very heavy users and shouldn't have a problem keeping a 3 year MTBF drive for more than three years (most of the time, there are outliers of course).

Of course, we will have at least some of these drives being made more reliable and having at least 5 year MTBFs.

All of this is assuming we get hard drives that utilize this or a similar technology. Laser writing hard drives might never see the light of day as widely used technology.

Your claim that they won't last long enough and won't be reliable enough seems invalid.
Score
0
!