[Beowulf] Is there really a need for Exascale?

[Eugen Leitl]

On Thu, Nov 29, 2012 at 04:57:15AM +0000, Lux, Jim (337C) wrote:

> But is that really true.   Sure, the processor in a cellphone is slower
> than say a typical modern PC CPU.. But, given appropriate software, is it
> a better $/FLOPS or W/FLOPS deal to get 100 cellphone CPUs or 1 superduper
> PC CPU?

The classical CPU core seems to be regressing to the role of a co-processor,
and real heavy lifting is being done with OpenCL, a la APU. Need to 
be able to process video and smooth game physics pushes a
GPU into every mobile device CPU. As it's memory starved, there's already
memory stacking http://www.ifixit.com/Teardown/Apple+A6+Teardown/10528/1?singlePage
I presume Exynos 5 teardown would not look much different.

The interesting part about forthcoming real (TSV) stacking 
that you can mix hybrid memories, whether SRAM, MRAM, DRAM, flash,
as you don't have to be process compatible as with true embedded 
memory (though MRAM might well be process compatible with CPUs).

Especially with static designs like MRAM power dissipation is low,
and MRAM will go where flash cannot, as it as the limit of its
scaling. If your CPU is also spintronics, then power dissipation
is only limited to when you need to flip spins, not pump electrons 
through leaky devices. Spintronics scales to single electron
devices, including single molecule scale (~1 nm).

http://arstechnica.com/science/2012/08/a-single-molecule-magnet-may-enable-quantum-computing/

Le Moore est mort, vive le Moore!