Still in the development stages but the technology discussed in the following report would seem to have great potential for silencing/low power storage:

http://www.theregister.co.uk/2009/06/15 ... erkeleydb/

No spinning disc, impressive performance and green credentials and I assume the 'oscillating' surface doesn't make much noise either. The question is, could it be produced economically enough to be viable?

As impressive as the "holographic DVD" or "flexible e-paper screen". In other words it looks good on paper and possible prototypes, but will never (or in the near future) come to the consumer market. Or to put it simply, I'm sick of these new "technologies"; when will we get to see something new in the local store?!

SSDs are more of a threat to this technology than they make out. Speed is constantly increasing, capacity exploding, prices crashing, power usage negligible etc. Good on them for trying something new, but they have no easy task ahead of them.

The only major problem SSD is facing is that a growing number of analysts are saying that the technological hurdles to go smaller than 22nm may be nearly insurmountable. Since 22nm is about a third the square size of today's 40nm flash memory, if we really did hit a wall at 22nm, that would mean drive capacities only about 3x higher than todays drives, with all else equal. The question would then become one of cost and if there's any other viable alternatives, or if getting a roadblock at 22nm would be a showstopper.

Eh. I seem to remember similar concerns about 45nm a few years ago and 100nm back in the days of the original Pentium (800nm). This is not unique to chips. Look at all the people at the dawn of the jet age who were sure that no human pilot and/or plane could survive exceeding the sound barrier. Before it was done, most experts thought a human could ever run faster than a 4min mile. In addition to it being very much in keeping with human nature to poo-poo all sorts of things outside of conventional thinking as impossible, there is also the fact that it is the job of analysts to be provocative. They make their living by selling their analysis, and there is not much of a market for saying: "everybody has everything figured out and nobody needs our advice."

Oh I'm not saying it can't be done, but right now Intel's "tick-tock" strategy is for roughly a doubling of chip density every 2 years. (i.e. 32nm has 2x as many transisters per square mm as 45nm, which is 2x of 65nm) There may be an extra year or two pause when we hit 22nm before they can drive the costs down enough on either extreme ultraviolet lithography, or xray lithography, or whatever it takes. Either that or we may soon see the transition away from silicon, and you can be sure that will cause delays in node reduction.

I remember reading an article about this once with an interview with intel. They said they see a clear path all the way down to 11nm without any major issues.

Well considering the only major technology Intel has been wrong about lately was Pentium 4's hitting 5ghz, other than that, they never miss a product delivery. I don't know what they have up their sleeve, but hey, if they can do it, great.