OT: This Swift thing

Gregory Ewing greg.ewing at canterbury.ac.nz
Mon Jun 16 18:12:13 EDT 2014


Anssi Saari wrote:
> That was before 90 nm when leakage current started dominating over
> switching current.

Well, if you don't care about speed, you probably don't
need to make it that small. There's plenty of time for
signals to propagate, so you can afford to spread the
circuitry out more.

The point is that "optimising for power consumption" on
its own doesn't really make sense, because there's no
optimum point -- you can more or less make the power
consumption as low as you want if you *really* don't
care about speed in the slightest.

In practice, people *do* care about speed, so it
becomes a tradeoff between low power consumption and
something fast enought that people will want to use
it.

> A few years ago jumbo sized but cheapish CULV laptops suddenly had 10
> hours plus battery but did anyone notice or care?

I think people do care, it's just that going from
something like 6 hours to 10 hours is not a big
enough change to warrant much hype. If it were
100 hours, without losing too much else, I'm
pretty sure it *would* be made a marketing point!

-- 
Greg



More information about the Python-list mailing list