bit-tech.net

IBM shrinks magnetic storage to just 12 atoms

IBM shrinks magnetic storage to just 12 atoms

IBM's prototype antiferromagnetic storage system measures just twelve atoms per bit - 100 times more dense than current systems.

Scientists working at IBM Research's Almaden facility have released details of a new breakthrough in magnetic storage: a memory bit just 12 atoms wide, around a hundred times the density of today's storage systems.

Using an unconventional form of magnetism known as 'antiferromagnetism,' the team of researchers have been able to create an experimental magnetic memory device twelve atoms in size; blowing past Moore's Law - the observation of component density doubling roughly every two years originally made by Intel co-founder Gordon Moore in a 1965 paper on integrated circuits - and paving the way for exascale storage systems in the future.

Unlike traditional magnetism, where nearby magnets are all polarised in the same direction, antiferromagnets alternate between north-polarisation and south-polarisation; as a result, it's possible to pack them significantly closer together than with traditional ferromagnets.

Offering a density some one hundred times greater than today's magnetic hard-drives or solid-state storage chips, IBM's breakthrough paves the way for future exascale storage systems with discs holding upwards of 200TB each.

Before you get too excited, however, there's a catch: the technology is far from simple to implement. IBM's prototype involves extremely low temperatures - around half a degree above absolute zero for maximum stability - and the use of an expensive scanning tunnelling microscope (STM) probe to flip the magnetic state of each antiferromagnet. As a result, it's unlikely we'll be seeing the technology hit the consumer market any time soon.

As a concept, however, IBM's work eclipses that of other teams working on similar projects; and, given time, could redefine storage as we know it.

Are you impressed to see just how small IBM can go, or will it take the technology appearing on the shelves of your local PC World before you get excited? Share your thoughts over in the forums.

14 Comments

Discuss in the forums Reply
Tattysnuc 13th January 2012, 15:38 Quote
Hardly a practical working concept is it. More a "we can do something really neat" statement than a proof of concept, especially given the fact that you are essentially talking about superconductor properties.
Gareth Halfacree 13th January 2012, 15:48 Quote
Quote:
Originally Posted by Tattysnuc
Hardly a practical working concept is it[...]
No, but neither was the first (bottle-sized) transistor. IBM reckons it works at room temperature, too, but needs 150 atoms instead of 12; that's still pretty damn small.
Guinevere 13th January 2012, 16:05 Quote
Since when was it claimed that Moore's Law applies to experimental lab based magnetic storage mechanisms? Moore's Law is about transistors (The general understanding anyway).

It's Kryder's Law that applies to hard drives http://bit.ly/xyD1sP so even saying this experiment pushes past said law is stretching things a bit far, seeing as it's 100% impractical to put into practice at the current point in time.

It's a lab trick, like 21 years ago when they wrote IBM in individual atoms http://bit.ly/xpQ1gj. Very clever and a show of what can be done in the lab, but not something mainstream.
Gareth Halfacree 13th January 2012, 16:07 Quote
Quote:
Originally Posted by Guinevere
Since when was it claimed that Moore's Law applies to experimental lab based magnetic storage mechanisms?
It was the comparative that IBM itself used, but you're right: Moore was originally talking transistors.
Tattysnuc 13th January 2012, 16:49 Quote
Quote:
Originally Posted by Gareth Halfacree
Quote:
Originally Posted by Tattysnuc
Hardly a practical working concept is it[...]
No, but neither was the first (bottle-sized) transistor. IBM reckons it works at room temperature, too, but needs 150 atoms instead of 12; that's still pretty damn small.

So at room temperature the storage density is approximately 10x that of current technologies (based on 12 atoms being 100x denser)

Wonder what the speeds of accessing data via this method is - it's almost pointless having all that storage space if it's slower to access than current technologies. Speed has been the biggest focus of late via SSD's, so I feel any technology would have to offer capacity as well as speed benefits. 200TB backup facilities would be fantastic, but it doesn't feel like a step forward for me - it's 2012 and the future of storage is little magnets? That doesn't feel like a new technology - more an optimization of one that's been done to death.

How robust a mechanism really is this?
PingCrosby 13th January 2012, 22:57 Quote
I built myself a compass using antiferromagnets before my trek across the Antarctica, got lost for months going round in circles and getting chased by bleedin polar bears. Bloody eBay.
Sloth 14th January 2012, 00:26 Quote
Quote:
Originally Posted by PingCrosby
I built myself a compass using antiferromagnets before my trek across the Antarctica, got lost for months going round in circles and getting chased by bleedin polar bears. Bloody eBay.
I think you were more lost than you originally thought, there are no polar bears in Antarctica.
alialias 14th January 2012, 00:31 Quote
^^ So much negativity! This is amazing, whether it's applicable to real life right now or not, the sizes they are talking about are ridiculous.
I will be posting this article to my Course Group on Facebook, and i'm sure some of them will be as excited about it as me.
As far as I can see too, this would mean faster access times too, surely if it's more dense, then there's less area to look for data? I don't really know about that sort of thing, so correct me if i'm wrong.
Phoenixlight 14th January 2012, 01:39 Quote
I wish they'd hurry up, I want a 1 petabyte 6GB/s HDD by next december!
DriftCarl 14th January 2012, 13:35 Quote
kinda need a 100TB USB stick nowadays.
ssj12 14th January 2012, 22:17 Quote
Quote:
Originally Posted by Tattysnuc
Hardly a practical working concept is it. More a "we can do something really neat" statement than a proof of concept, especially given the fact that you are essentially talking about superconductor properties.

This is IBM we are talking about. If they develop something, they use the tech.
LordPyrinc 15th January 2012, 05:13 Quote
Even, if this particular design runs to a commercial dead-end, perhaps some of the concepts and techniques will still provide value in other designs.
mclean007 15th January 2012, 13:57 Quote
Quote:
Originally Posted by Sloth
Quote:
Originally Posted by PingCrosby
I built myself a compass using antiferromagnets before my trek across the Antarctica, got lost for months going round in circles and getting chased by bleedin polar bears. Bloody eBay.
I think you were more lost than you originally thought, there are no polar bears in Antarctica.
I think that was the joke - according to the article "antiferromagnets alternate between north-polarisation and south-polarisation" so PingCrosby's home made compass took him to the north pole instead of south.
PingCrosby 16th January 2012, 12:26 Quote
Quote:
Originally Posted by mclean007
Quote:
Originally Posted by Sloth
Quote:
Originally Posted by PingCrosby
I built myself a compass using antiferromagnets before my trek across the Antarctica, got lost for months going round in circles and getting chased by bleedin polar bears. Bloody eBay.
I think you were more lost than you originally thought, there are no polar bears in Antarctica.
I think that was the joke - according to the article "antiferromagnets alternate between north-polarisation and south-polarisation" so PingCrosby's home made compass took him to the north pole instead of south.

Thank god someone else apart from my psychiatrist understands me.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums