Slashdot Mirror
'Millipede' Prototype Shown at CeBIT
neutron_p writes "It was a subject of much controversy for last 5 - 7 years, but it's finally got protyped. At CeBIT, IBM for the first time shows the prototype of "Millipede" - nanomechanical data storage device. Using revolutionary nanotechnology, scientists at the IBM Zurich R&D Lab, Switzerland, have made it to the millionths of a millimeter range, achieving data storage densities of more than one terabit per square inch, equivalent to storing the content of 25 DVDs on an area the size of a postage stamp. The principle of operation is comparable with the old punch cards, but now with structural dimensions in the nanometer scale and the ability to erase data and rewrite the medium."
1. What's the read/write speed?
2. What's the operating temperature requirements?
3. What's the max operating heat output per unit?
4. How many concurrant inputs/outputs can we get into a unit?
5. What's the failure rate/expected operating lifespan?
6. What's the near-term expected commodity cost of these units?
7. Given 1-6, how many units would be needed to make a properly redundant filesystem with at least the reliability and speed of current file storage devices on the market? What would be the expected near-term cost?
Ryan Fenton
For those interested, here are some advantages I see to this technology:
1. Increased storage density. More importantly, this prototype is not near any fundamental limit. Hence, it would appear that there is plenty of room to reduce the dimensions of the MEMS tips to increase storage densities way past what a magnetic drive can do.
2. Data transfer rate. In principle, the thousdands of different tips can all return data at the same time, compared to, say, 4 bits returned at once from a 4-platter HDD. Of course, in real situations, not all 4000 bits will necessarily be of interest, but I think with smart caching and device layout the throughput should be very high (i.e.: contiguous bits in a file are spread out so that the entire file is read by the 4000 tips without anything moving).
3. Low seek times. In a HDD, the head must move by many centimeters in order to seek randomly. In Millipede, the entire surface moves by, at most, 100 micrometers to find a new location. It probably uses piezoelectrics, which are fast and robust. Thus, I see seek times being lower (at least in a mature device).
4. Scalable. This prototype has a single array of tips on a single polymer layer. Obviously it is straightforward to build real devices using 10 or 20 of these arrays stacked. Unlike the platters in a HDD, these arrays could be seeking independantly, so if properly designed, performance could be very good (like RAID maybe?).
5. Heat. The piezos shouldn't heat up too much, and even though the tips themselves use pinpoint heating to deform the polymer, I think the bulk device heat would be lower than a HDD spinning at 10k rpm. Less noise too.
6. Cost. By using established MEMS technology (i.e.: the same lithography used to make microchips nowadays) I don't think implementation costs (and future scaling) will be too expensive (as compared to some more far-fetched nanotech ideas).
This has been in the works for a long time, but I think we may actually see real devices soon! (6 years?) I think this technology has real potential, and I think IBM is right to pursue it.
http://en.wikipedia.org/wiki/Compact_disk
:)
The information on a standard CD is encoded as a spiral track of pits moulded into the top of the polycarbonate layer
Sometimes it's true: the more things change, the more they stay the same. The preffered method for lengthy data storage still involves making an impression.
The oldest methods of "data storage" go back to the birth of written language. These involved either making impressions in the sand, or for more permanent storage making engravings into stone.
How small our stones have gotten, eh?
This technology sounds wonderful, but at the same time dangerous. You could have some extremely potent monitoring devices with this thing: Cameras that can record for weeks or months, microphones can record for years, etc. Then again, the practical uses sound great, our ram is going to be forever changed, and I wont need to sweat over 1gb.
We're getting there, but we're not there yet. And we won't be until storage is truly ubiquitous. I've actually spent some of my weekend re-organizing my music collection, ripping CDs that hadn't listened in a while, etc. But even with the 600G of storage in my PC, I still can't have everything I want unless it's compressed. And I'm thinking about how to listen to my collection in my car. Bringing hundreds of CDs around with me isn't practical. MP3 CDs hold maybe 10-20 albums. HDD based devices (ipods and the like) still can't hold everything I own... not even close. And I want to have a DVD server so rather than pulling out the DVD, I can just call up one of the hundreds of DVDs I own on a menu.
Yes, storage is becoming more impressive all the time. But it's still a very long way from being to the point where you don't have to think about how and where you store and move your files. And it will be very cool when that day comes.
-S
--- What parts of "shall make no law", "shall not be infringed", and "shall not be violated" don't you understand?
I'm really glad that there are still American companies around that are doing fundamental technological research that will improve our lives in the future. Sure IBM may be huge and somewhat evil in it's own way, but at least they know how to actually invent useful things, rather relying on lawsuits and dubious claims of "intellectual property" and whatnot to extract wealth from others.
We're all going to be out of work in a few years if this continues! /sarcasm I really like advances like this because it saves us time. Imagine what politics would look like if all of the IT brains that are writing redundant perl scrips suddenly applied their brains to history and politics. It'd probably change the world.
It's just like the industrial age, we can put down our sledge hammers(mice) and redirect our energy to more important things.
What if Digg added local news and a Slashdot inspired comment karma system? ---
http://houndwire.com
That's why you punch sequence numbers in columns 73-80 - so you can put them back in order. Writing a diagonal stripe across the edge of the deck helps, too.
Fiat Lux.
It seems to me that this kind of technology has been IBM's wild card for a long time. I think they've got a very good idea of what the face of the computer world will look like in a couple years, and they're doing everything they can to come out ahead. First they become a linux house, most likely because linux has proved to be a very nice archetecture to do things like clustering. Now they're finally using the nanotechnology they've been working on for years in such a way that they've created an amazing new technology like this. A technology, I might add, that has the potential to completely dominate the market and completely change the face of the computer world to the point where IBM is the largest hardware manufacturer in the world.....yet again. I'd love to see what's in their business plan for the next few years.
Total device: 6.4 mm length, tip pitch 100 um
/sec
-> 64 rows and 64 columns
-> 4096 tips
Writing speed (from TFwebsite): 'a few microsecond' (say 10)
-> 4096/10e-6 = 410 Mbit
Per tip: range 100 um, bit pitch 10 nm
-> 10000 x 10000 bits = 100 Mbit
Position resolution (really neat device using micro-heaters): 2 nm over 120 um ->
-> 60000 positions observable (probably 16 bit)
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
The question behind the questions is what potential roles that this product could fill.
If it can't run at room temperature conveniently, but can be made cheap per storage space and is reliable, then it may be useful in stationary servers for extreme-mass remote storage.
If it can run at room temperature and is somewhat affordable, but slow, it can be used as common backup.
If it can end up close but superior to hard disk in all aspects, then it may replace them.
If it can be fast enough to be used as live memory at room temperature, with conventional memory as cache, then even with a few limitations, it could transform the nature of computers as we experience them.
There's many, many other possibilities. Yes, of course, as you suggest, price will match the market - but the role this technology can play is limited more by it's logical capability than the market. If the possibility is open, it's usually much more of an opportunity if you can create a new technology in a market than to just replace another. That's why my questions are obvious - we all wonder how far this first generation of nanotechnology will take us.
Ryan Fenton
He's not joking, he's expressing disgust with the media's continuous habit of dumbing down units of measurement to the point where they're meaningless. It's irritating to those who actually know what a square millimeter and a bit are.
If a job's not worth doing, it's not worth doing right.
Not to take away from the extreme coolness of this, since it is cool, but it's not nanotechnology. It's built using microelectronic fabrication techniques. We're a long way from nanofabrication yet.