Secondly, after double-checking the abstract, I made the editorial decision to replace the subtitle "nanoarchitectonic design of a complex system for natural computing" with "design of a functional device". My reasoning was that the original subtitle seemed hopelessly opaque (outside the field, presumably), and that the development of a testable device seemed to be the most interesting news.
Thirdly, I know next to nothing about the topic, and am only guessing that this is newsworthy. My introduction to the news was from a post at phys.org by Lisa Zyga (http://phys.org/news/2015-05-scientists-atomic-scale-hardwar...). Phys.org is a frequently breathy science news site, but Ms. Zyga has been blogging there for years and is a decent science reporter.
I am hoping that this post will be of some interest to those with domain knowledge, and that I can learn from any discussion or criticism of the paper.
I did research with Adam & Jim back in undergrad @ UCLA - I had a first stab attempt at creating these atomic switches & failed miserably. I'll see if I can get them to comment if there's a specific question here.
clever, wish we could read the research. Seems that the aim is to create a higher density (and presumably larger scale) medium for computing than can be made with traditional fabrication. trade-off (opportunity?) is that it computes much differently - you learn to compute with the switches you get, rather than building the switches you need. probably has some nasty sensitivities to voltage, and RF, and heat, and vibration - but that's just engineering :)
Have they demonstrated much better density yet? (from skimming one of the researchers' other papers they have gotten 10^8 switches/cm^2 which is not as dense as the transistors on a high-end graphics card. not that they are really directly comparable anyhow, this would be like a special coprocessor for neural-stuff.
3 comments
[ 3.3 ms ] story [ 19.0 ms ] threadFirstly, the paper is paywalled, but there is an article discussing it at http://phys.org/news/2015-05-scientists-atomic-scale-hardwar....
Secondly, after double-checking the abstract, I made the editorial decision to replace the subtitle "nanoarchitectonic design of a complex system for natural computing" with "design of a functional device". My reasoning was that the original subtitle seemed hopelessly opaque (outside the field, presumably), and that the development of a testable device seemed to be the most interesting news.
Thirdly, I know next to nothing about the topic, and am only guessing that this is newsworthy. My introduction to the news was from a post at phys.org by Lisa Zyga (http://phys.org/news/2015-05-scientists-atomic-scale-hardwar...). Phys.org is a frequently breathy science news site, but Ms. Zyga has been blogging there for years and is a decent science reporter.
I am hoping that this post will be of some interest to those with domain knowledge, and that I can learn from any discussion or criticism of the paper.
Have they demonstrated much better density yet? (from skimming one of the researchers' other papers they have gotten 10^8 switches/cm^2 which is not as dense as the transistors on a high-end graphics card. not that they are really directly comparable anyhow, this would be like a special coprocessor for neural-stuff.
here's a good overview with quotes from one of the above paper's authors: http://nanotechweb.org/cws/article/tech/61049