News - Data storage miniaturisation; how about one bit per molecule?

[HEXUS]

Using nanoscale molecular switches.

Read more.

[cameronlite]

Did anybody else pick up on the fact that they are saying 1 bit per molecule and the current tech is on 1 bit per 3 million atoms. It's not a like for like comparison...

[miniyazz]

51 atoms per molecule apparently.

[aidanjt]

Of course the problem with nano-scale storage is the data is much more readily corruptible. Genetic coding had to evolutionarily bulk up to protect itself for a very good reason.

[Hicks12]

yup indeed, im not really sure how to take that comparison... its completely wrong . If it was 1 bit per atom then id say, great stuff

[b0redom]

Obligatory physics joke:

Two atoms walking down the road. One says, "Wait I've dropped an electron we need to go back."
The other replies, "Are you sure?"
The first answers, "Yes. I'm positive."

[capt_cornflake]

I suspect it may be quite some time before an STM becomes a component of common desktop systems.

[crossy]

I'm cynical today - surely this is the latest in a long line of data storage "perpetual motion machines"?

I remember IBM claiming years ago that we'd all be storing Terabytes of data in something the size of a sugar cube - optically I think - by 2015. Can't see that prediction coming to pass.

So yes, from the physics point of view this latest achievement is very impressive, but I'll be more impressed when they've got something that can RELIABLY store something approaching the massive amounts of data promised.

That said, the growth in data storage is fascinating - I came across a 3.5" Seagate 40MB drive at the weekend, (ex Amiga boot drive), and sitting on my mouse pad at the moment is an old uSD card - 6,000MB.

[DeludedGuy]

Heh, I remember getting a 10GB drive and thinking I would never fill it up...

[pipTheGeek]

I remember being totally in awe of a 40mb SCSI HDD at school. Then being amazed when the first 120MB drive became available for my Atari ST. Not that I could ever afford one.

Some years ago I asked my geek friends how long they thought it would be before these new fangled flash cards / sticks replaced CDs as the re-writable media of choice. The consensus was that the flash memory was never going to get cheap enough to displace CDs. Now companies give away 2gb memory sticks like they used to give away pens.

Back on topic, what every happened to carbon nano-tube RAM? Wasn't that meant to be mainstream by now?

[MSIC]

1 bit per molecule? Hah. Old news.
I think we should be aiming to use Mr Higg's very own Boson now....

[warejon9]

Of course the problem with nano-scale storage is the data is much more readily corruptible. Genetic coding had to evolutionarily bulk up to protect itself for a very good reason.

When you say bulk, yes there is some spare stuff that doesn't encode for anything, however it is important for the evolution of proteins, allowing proteins to be relatively modular. Also when you think that your DNA effectively codes for everything about you biologically for however long you live, its not too bad, and it does it using a simple 3 part rule.

DNA is interesting as it has its own inbuilt repair system (the base pairing) which only goes wrong when a lot of single nucleotide polymorphisms happen.

[miniyazz]

If it used RAID 5 it would be a lot more efficient

[crossy]

If it used RAID 5 it would be a lot more efficient

At the risk of getting yelled at as an idiot by the folks that know more about this than I do - which, let's be honest, is just about everyone else - here's a question.

If the current tech stores 1 bit in 3 million atoms and this new one stores it in 51, then surely that gives enough "spare" space to put in some staggeringly impressive ECC and still demonstrate some pretty impressive density increases. E.g. use 51 atoms to store one bit, use another 400,000 to store error correct/parity information and still get a six-fold increase in data density.

Apologies if this is a "Sun reader" level of stupidity, but I've only got dimly remembered high school physics to fall back on.

[howdee]

If the current tech stores 1 bit in 3 million atoms and this new one stores it in 51, then surely that gives enough "spare" space to put in some staggeringly impressive ECC and still demonstrate some pretty impressive density increases. E.g. use 51 atoms to store one bit, use another 400,000 to store error correct/parity information and still get a six-fold increase in data density.

We haven't heard anything about the molecular density or indeed the distance between each molecule required to prevent them from involuntarily switching states and - to keep it short - we don't have nearly enough data to support any such claims either way. We're left with having to trust the researchers it's indeed a data density breakthrough and they wouldn't have published their findings if it weren't so. As for your figures - my guesstimate would be you were roughly 6.000 times humbler reading this news article than I was.

[watercooled]

Misread FTW.