Moore’s Law Not Dead – Intel’s Use of HPC to Keep It Alive

Re:Moore’s Law Not Dead – Intel’s Use of HPC to Keep It Alive by jmgrosen on Tuesday January 12, 2016 @11:09AM (#10888163)

Or maybe Intel really has hit a thermal limit, rather than hatching a conspiracy against consumers?

(Not saying 10+GHz is necessarily impossible, just that it's more likely hitting it is hard than Intel doesn't want to innovate there for some reason.)

[ Parent | on HN ]

Re:Moore’s Law Not Dead – Intel’s Use of HPC to Keep It Alive by frik on Tuesday January 12, 2016 @11:22AM (#10888247)

AMD sells CPUs with 4.2+ GHz that can be easily overclocked (like Intel CPUs from 15+ years ago). We are now at 12nm whereas we had >90nm years ago. Increasing the clock speed is certainly possible, but always a tradeoff. We can always put the CPU in a specific oil like the Cray2. If x86/64 is the limiting factor, they could release an incompatible CPU architecture (like the failed Itanium or like ARM). A lot of the die space on current CPUs is internal memory (L1, L2). It's pretty easy to draw the 'conspiracy' card.

[ Parent | on HN ]

Re:Moore’s Law Not Dead – Intel’s Use of HPC to Keep It Alive by PeCaN on Tuesday January 12, 2016 @05:09PM (#10890797)

AMD's CPUs do many fewer instructions per cycle as Haswell (in some cases half as many), not to mention Skylake. You can't really compare GHz between them.

And releasing a new architecture is not exactly feasible if they want to stay in the server, workstation, and desktop markets. That idea behind Itanium actually was to get faster by dropping x86_64, and we see how poorly it worked out.

[ Parent | on HN ]