[HH] The Death of Moore's Law Will Spur Innovation

[Kurt Keville]

It is fortunate bunnie constrained his arguments to "classical" Moore's Law
since if he had expostulated a little more on the various "goodness"
variables he would find considerably less agreement there. For one of the
more psychedelic takes on "extended" Moore's Law, read Kurzweil's "The
Singularity is Near"

That having been said, Ron Luijten says, at panels like the one below, that
Intel may be at a node size too small already and that when voltage leak
equals useful voltage you have to look at a different business model, not a
different chemistry...
http://sc14.supercomputing.org/schedule/event_detail?evid=pan122

Ron also says the Von Neumann architecture is dead. I will describe his
approach at my lecture Wednesday at BLU... and if you would rather get into
his face rather than mine he is coming to MIT in July.
http://majorca-mit.org/

Kurt


On Fri, Apr 10, 2015 at 3:37 PM, Tom Metro <tmetro+hhacking at gmail.com>
wrote:

> Here's an opinion piece by Andrew "bunnie" Huang (hardware lead at
> Chumby, and I believe the guy I previously posted about who was building
> an open hardware laptop design).
>
> The Death of Moore's Law Will Spur Innovation
>
> http://spectrum.ieee.org/semiconductors/design/the-death-of-moores-law-will-spur-innovation
>
> His assumption is that when semiconductors approach an "effective gate
> length of about 5 nanometers sometime around 2020 or 2030", Moore's Law
> will hit a wall and the doubling of transistor density will drop from 18
> months to 36 months or more.
>
> He argues that the current reality of 18 month time horizons to
> substantially better silicon means that all the innovation is happening
> at the large organizations designing and producing the silicon. All the
> smaller companies making use of the silicon are better off waiting for
> the next CPU advance, rather than investing effort into code
> optimization and other performance optimizations that are within their
> grasp.
>
> But once the pace of big advances in silicon slows to something close to
> 36 months, it then makes sense for small organization to invest in such
> optimizations.
>
> Furthermore, he believes because we will be using the same silicon for
> longer periods, that will lead to more standardized components (think
> Google's modular Aria phone, or PC motherboards that aren't obsolete in
> 9 months), computers and electronics will be better built to last
> longer, and we'll go back to a "repair culture" where schematics and
> replacement parts are readily available.
>
>
> This all sounds great for those rooting for open hardware, but all of
> this flows from the starting assumption that Moore's Law will run out of
> gas. The problem with that assumption, even though it is supported by
> the laws of physics, is that there will be many highly motivated
> organizations with deep pockets that will seek to redefine the problem.
>
>  -Tom
>
> _______________________________________________
> Hardwarehacking mailing list
> Hardwarehacking at blu.org
> http://lists.blu.org/mailman/listinfo/hardwarehacking
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.blu.org/pipermail/hardwarehacking/attachments/20150410/a6b795d1/attachment.html>