Actually it could be much better than one bit per atom.
It depends on how accurately the distance between atoms can be measured and how closely they can be packed. On a 2D surface there might be an analogy to linear RLL codes. Current digital storage systems use "Run length limited" (RLL) codes where it's the distance between bits that contain the information. Theoretically you only need two samples to get lots of bits - if you can measure the distance to an accuracy of 1:4096 you get twelve bits per atom! ( Last I knew, hard disks use RLL 2,7 meaning minimum space is 2 units and max is 7 units, and CDs use RLL 8,14.) From: Jed Rothwell [mailto:jedrothw...@gmail.com] Sent: Thursday, May 2, 2013 7:51 AM To: vortex-l@eskimo.com Subject: Re: [Vo]:IBM Stop Motion Film of Cu Atoms Come to think of it, this means that in principle IBM could store data at one bit per atom starting now. Perhaps the biggest difficulty would be finding the data again. I guess this is the lower limit to data storage. I doubt that subatomic storage will ever be possible....
