Yes that's more of less what SA said - they've got around the clock speed limit by multiplying cores, but they can't get around the fact that components can't be scaled below (I think) 14nm without that transistors leaking electrons - at least not without some radical new technology. So it was about whether some new paradigm is on the way to keep things heading towards the what's his name - begins with L I think's - limit. (Memristors nanotubes etc)
Anyway, have you read the article? You can probably make more intelligent comments on it than me. PS Laudauer?!? On 27 April 2015 at 14:23, Russell Standish <[email protected]> wrote: > On Mon, Apr 27, 2015 at 01:48:40PM +1200, LizR wrote: > > According to the latest Scientific American, Moore's Law stopped working > > about 10 years ago. I'm not sure if or how this affects the > > prognostications for AIs, mind simulation etc, though. > > The only thing that stopped 10 years ago was the increase in CPU clock > speed. > > That was never Moore's law, though, which refers to density of > transistors for a given price point. That has very much continued to > increase. I can now buy CPUs with 50 cores for the price of a dual > core system 10 years ago. And each core has almost an order of > magnitude performance improvement due to architectural improvements > (eg more cache, hyperthreading, SIMD/vector instructions etc). That's > about 200 x performance improvement over a decade, about double what > Moore's would predict. But its all parallel computing - its not going > the make Microsoft Word any less of a dog. > > Cheers > > -- > > > ---------------------------------------------------------------------------- > Prof Russell Standish Phone 0425 253119 (mobile) > Principal, High Performance Coders > Visiting Professor of Mathematics [email protected] > University of New South Wales http://www.hpcoders.com.au > > ---------------------------------------------------------------------------- > > -- > You received this message because you are subscribed to the Google Groups > "Everything List" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/everything-list. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/d/optout.
