Groundbreaking 'atomic memory' could cram unimaginable amounts of data into your iPhone

Posted:
in Future Apple Hardware
A new technology for data storage at the atomic level could supplant magnetic and flash media in the future, and yield thousands of terabytes in a single drive -- if scientists can perfect it.




Scientists have demonstrated the use of the location of single atoms to store bits of information in a study published by scientific journal Nature Nanotechnology. The new technology, dubbed atomic-scale memory, used a copper bed and 60,000 chlorine atoms, scattered at random and has managed to retain more than a single bit of information at the time.

Storage is measurable in areal density, or a measure of the quantity of information that can be stored in an area of media. The researchers claim that the areal density of the technology is 502 terabits per square inch. For comparison, currently shipping hard drives peak at an areal density of 1.34 terabits per square inch. Dual-layer Blu-ray disks peak at around 60 gigabits per square inch.

At present, the technology uses a scanning tunneling microscope to move and read the atoms' position. The relationship between an atom's location, and empty space on the copper bed encodes the data, as well as provides a primer for how to read the grid of atoms.

The bed first stored a part of Charles Darwin's "On the Origin of Species." After an initial write, the memory was then replaced with 160 words from a lecture by physicist Richard Feynman from 1959 that is credited for the first theories about atomic-scale data storage.



Feynman's lecture stored on the copper bed. Dark blue dots are single chlorine atoms.


The initial write, and copper bed seeding, took about a week for a single kilobyte of information. The rewrite of the partially organized copper plate took a few hours.

Besides the drawbacks of formatting and write speed measurable in hours or days, the process also must be maintained at temperatures as close to absolute zero as possible. The data was stable up to 77 K (-321 F), which is much higher than previous experiments in the field. However, without maintaining the extremely low temperature, the normal migration of atoms destroys the stored data.

The researchers are moving on to other materials with physical properties similar to the chlorine and copper array, in an attempt to stabilize the lattice at higher temperatures.

"It's automated, so it's 10 times faster than previous examples," said IBM staff scientist Christopher Lutz about the breakthrough. Lutz also noted that "it's important not to see this one-kilobyte memory result as something that can be taken directly to a product."
«13

Comments

  • Reply 1 of 42
    thrangthrang Posts: 755member
    This sounds like something Samsung should copy right away and launch in the fall...
    MacsAlwayscalirepressthis
  • Reply 2 of 42
    gatorguygatorguy Posts: 19,805member
    thrang said:
    This sounds like something Samsung should copy right away and launch in the fall...
    ...with a free pair of winter gloves for pre-orderers. 
    jbdragonMacsAlwaysrevenantlatifbprepressthis
  • Reply 3 of 42
    maestro64maestro64 Posts: 4,414member
    I love reading this stuff, then you get to the gotcha, you have to be at almost absolute zero for it to work. Another example of interesting experiment but it will never see the light of day as a product. At least IBM is still involved in fundamental research like this. IBM has made most of the breakthroughs in storage technologies, most of which never yielded a product. Look up Raceway Memory.
    bdkennedy1002jbdragonrevenantSpamSandwich
  • Reply 4 of 42
    Lost me at "the process also must be maintained at temperatures as close to absolute zero as possible."
    jbdragonSpamSandwich
  • Reply 5 of 42
    harrieharrie Posts: 1member
    Why are you crediting IBM when the research was from Delft University?
    René de Katirelandrepressthis
  • Reply 6 of 42
    ceek74ceek74 Posts: 323member
    "could"
  • Reply 7 of 42
    cnocbuicnocbui Posts: 3,613member
    "Groundbreaking 'atomic memory' could cram unimaginable amounts of data into your iPhone" If it was the size of a house.
  • Reply 8 of 42
    nolamacguynolamacguy Posts: 4,758member
    maestro64 said:
    Another example of interesting experiment but it will never see the light of day as a product.
    yes, because experimental research *never* evolves into real solutions for real problems in real products through years of hard work and iterative improvement. 

    smh. do you guys even listen to yourselves? you do realize that practically any modern computing tech was once an impractical experiment? as if flash memory just hopped out of a seashell, fully formed and ready to slap into your *handheld computer*. have you seen the size of the first storage devices?!? they were dresser-size. 
    edited July 2016 calimike1williamlondonafrodriButidonttweetfastasleeprepressthis
  • Reply 9 of 42
    MarkTwoMarkTwo Posts: 1member
    harrie said:
    Why are you crediting IBM when the research was from Delft University?
    +1

    Dutch/Portugese/Spanish research team. What did IBM have to do with this?
    irelandrepressthis
  • Reply 10 of 42
    cnocbuicnocbui Posts: 3,613member
    harrie said:
    Why are you crediting IBM when the research was from Delft University?

    Probably because IBM deserves the credit as they invented the technique and the equipment that makes it possible.  This research is iterative.

    Twenty years ago this week, on Sept. 28, 1989, an IBM physicist, Don Eigler, became the first person to manipulate and position individual atoms. Less than two months later, he arranged 35 Xenon atoms to spell out the letters IBM. Writing those three characters took about 22 hours. Today, the process would take about 15 minutes.



    edited July 2016 DeelronMacsAlwaysrepressthis
  • Reply 11 of 42
    tallest skiltallest skil Posts: 43,399member
    It’ll never come to production.
    SpamSandwich
  • Reply 12 of 42
    Finally, a reason to put on my Mr Freeze suit. 
    ireland
  • Reply 13 of 42
    mtbnutmtbnut Posts: 190member
    thrang said:
    This sounds like something Samsung should copy right away and launch in the fall...
    You have a point, as the performance lights up Samsung's spec sheets:
    The initial write, and copper bed seeding, took about a week for a single kilobyte of information. The rewrite of the partially organized copper plate took a few hours.
    edited July 2016
  • Reply 14 of 42
    jbdragonjbdragon Posts: 1,964member
    cnocbui said:

    Probably because IBM deserves the credit as they invented the technique and the equipment that makes it possible.  This research is iterative.



    That really is pretty cool.
  • Reply 15 of 42
    It’ll never come to production.
    Are you sure about that?  Never is a long, long time.
    mike1nolamacguycaliafrodrifastasleepjony0repressthis
  • Reply 16 of 42
    tallest skiltallest skil Posts: 43,399member
    Are you sure about that?  Never is a long, long time.
    For the same reason I’m certain that we won’t ever have pneumatic tubes to the home as was discussed in the 1920s, yes. Some technologies only exist to expand the capabilities of others.
    revenantSpamSandwich
  • Reply 17 of 42
    evilutionevilution Posts: 1,342member
    It’ll never come to production.
    I have to agree. If they ever manage to work around the temperature problems, the setup speed, the read and write speed, it'll start off incredibly expensive and will be so far down the line that other technologies will have surpassed it on price and speed.
    DNA drives, holographic discs and further advances in HDD and SSD will render atomic scale memory not worth pursuing.
    SpamSandwich
  • Reply 18 of 42
    This will be expensive considering how Apple charges $100 for a 16GB upgrade.
    tallest skilcnocbui
  • Reply 19 of 42
    mac_dogmac_dog Posts: 639member
    thrang said:
    This sounds like something Samsung should copy right away and launch in the fall...
    Oh, they'll fuck that one up, too.
    cali
  • Reply 20 of 42
    19831983 Posts: 1,142member
    This ain't coming to an iPhone or Mac for a very-very long time!
Sign In or Register to comment.