Slashdot Mirror
Stanford's Quantum Hologram Sets Storage Record
eldavojohn writes "It's often assumed that representing data reaches a limit when you get to the point that an atom represents one bit in some form or fashion. But Stanford University researchers have used a quantum hologram model to store the characters 'S' and 'U' by encoding the data at a rate of 35 bits per electron."
One thing most 'futurists' agree on is that the ultimate 'end game' of technology appears to be the conversion of all matter in the solar system into machine parts and computational elements. It's a logical end result of exponential growth. (and, actually, would be only the beginning : such a 'civilization' would eventually grow to convert the entire universe, but this would take much longer due to the snails pace of light)
It's neat to think that such a civilization could store even more information than an obvious cap of '1 bit per atom'.
If I understand holography and what they're doing correctly (and I DID work as a tech in Emmett Leith's lab so I have some clue), they're transforming the information.
Yes, each electron has information from 35 bits. But more than one electron has that same information, encoded differently. How many storage electrons do they need to encode it in a way that is recoverable?
The information per electron is the total information encoded divided by the total number of electrons needed to encode it at a high enough resolution to be recovered.
Also: The illustration of the way they're encoding it looks like it's not just electrons that encode it, but also their absence. Add in HOLES to the count of "things encoding the bits".
I'll be surprised if the total comes out to more than one bit per electron site. (Note that they may get more than one such site per atom.)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Why would an atom even naively be the conceptual finest representation of a bit? Does he mean an electron? An atom is quite a large and complex object compared to an electron...Measuring an electron's presence or absence yields a simple bit. There's nothing atomic ABOUT an atom... why would that ever represent a bit?
It sounds like cheating to me as well. They don't seem to be counting the MOLECULES necessary for creating the interference patterns. How many support atoms does it take to encode each bit of information? If it takes more than a couple for each bit, then how is this better than IBM's effort?
From the article:
On the two-dimensional surface of the copper, electrons zip around, behaving as both particles and waves, bouncing off the carbon monoxide molecules the way ripples in a shallow pond might interact with stones placed in the water. The ever-moving waves interact with the molecules and with each other to form standing "interference patterns" that vary with the placement of the molecules.
By altering the arrangement of the molecules, the researchers can create different waveforms, effectively encoding information for later retrieval. To encode and read out the data at unprecedented density, the scientists have devised a new technology, Electronic Quantum Holography.
If they can't recover the data, how did they prove it was ever there? I didn't read the article in good /. fashion, but if it avoids this question I'm sure it's not to be taken seriously.
Show a man some news, distract him for an hour. Show a man some mod points, distract him for the rest of his life.
I remember discussing related "small-scale storage" issues with my brother once. Two concepts were of particular interest:
1. Spin and such: If we want to store on a very small scale, why not use the intrinsic properties of molecules, atoms and particles? A simple example would be using a caffeine molecule, which can exist in 8 different molecular arrangements (I forget the exact details - was it aggregate Spin?), as 3-bit memory. I'm sure there are more suitable molecules, or applications on smaller scales, but the concept is sound.
2. Holographic storage: When part of a holographic surface is destroyed or decayed, it does not result in the hologram missing parts, but in a degradation of its overall clarity, since each area of the surface encodes a little of the information about the whole hologram. If storage could be designed around the same concept, data would not be lost unless enough of the whole holograph were destroyed or corrupted.
I particularly like this last idea, but unfortunately I suspect it would only work as permanent - not active - storage, such as read-only media. I think you could only write each bit of the hologram (or equivalent) if you knew what the whole was going to look like.
This study seems to demonstrate the same conceptual problem, although it isn't mentioned. The resultant "picture" could only be constructed by moving atoms around until the interference between their electrons produced the desired pattern. Trying to add to it would require a re-arrangement of the whole structure every time, and such arrangements would increase in complexity exponentially.
Meta will eat itself
But how long will the electrons stay in these different levels?