Recent developments seem to support the idea that computers are taking on a new character. It seems reasonable to consider alternate architectures that include memristors and FPNI. I see that there is a great advantage to be had in taking the CISC out of computation. There are memory architectures that would take advantage of the new methods that seem to be heading toward diskless operating systems. I would say that Linux and open source are positioned to get the greatest advantage from such changes. In linux, everything is a file and when everything gets flattened, it speeds up dramatically. The time overhead of disk is more than just access time.
It would seem that sol gel for oxides could easily be implemented by individuals. I am going to determine the methods and material necessary to create a circuit on glass or plastic. It seems to me that a different type of cpu would make more sense as digital is a precise thing, but becomes imprecise very quickly when faced with infinities. Perhaps good computation only needs to be a reasonably close guess.
That was a good laugh. I designed a memristor and a memistor 25 years ago and this is no different. It can be implemented with many different materials and what is happening at the molecular level is very interesting. There is no reason for me to fabricate this stuff as I already have a material system that is much better in many ways. I never saw this research, but I think it goes way back and intersects in a very strange way.
0 comments:
Post a Comment