Re: Moore's law resuccitated

On Jan 28, 12:50 pm, "Dragontamer" <prtig...@xxxxxxxxx> wrote:
On Jan 27, 11:15 am, "santosh" <santosh....@xxxxxxxxx> wrote:

http://news.bbc.co.uk/2/hi/technology/6299147.stm

It looks like Moore's law is going to be kept alive for some more
years.

When did Moore's law change into "double the amount of transistors"?
Last time I remembered it, it was double speed every year and a
half...

As others have pointed out, it has nothing to do with speed.

I would point out that his original paper said one year and he revised
this several years later to say two years. The "year and a half" value
was created by the media by averaging the two values. If you look at
the graphs, it has been following the "double every two years."

The other important thing to note is that doubling doesn't occur by
simply shrinking the transistors. This is also achieved by developing
processes to handle larger wafers (and die). Chips today are *much*
larger than they were back then, which is one of the reasons (along
with shrinking transistor sizes) that we've packed so many transistors
on a single chip.

It is also worth pointing out that "wafer-level circuits" haven't ever
become commercially viable. There are still too many defects on a
given slab of semiconductor material to pull this off.

Now as for the correlation between speed and Moore's law: in the
1990s, speed was beating Moore's law. In the first half of the 2000s,
speed was stalled (at least, when looking at the x86 and the bad path
the PIV took). Today, most speed gains are occuring by putting *lots*
of cache on the CPU. But that's not going to work forever. Bottom line
is that we'll probably soon see that the correlation between speed and
Moore's law bottoms out for uniprocessor CPUs. Speed doubling in the
future is probably going to occur via multiple cores. That, of
course, is not going to automatically equate to faster running
programs.
Cheers,
Randy Hyde



.

Relevant Pages

  • Re: Print to SHR
    ... Moore's Law has nothing to do with clock speed. ... transistor density into increased throughput per dollars (or cost- ... one thing that a surfeit of transistors has done is make ... increased single-thread parallelism. ...
    (comp.sys.apple2)
  • Re: Pete Lefferts LED current source
    ... current error) with unmatched transistors running at very different current densities, which shows that BJTs follow the Ebers-Moll law very closely, until the extrinsic resistances become important. ... Ebers-Moll is good to a couple of parts in 10**4 or better for small signal transistors if the CB bias is fixed, the transistor isn't saturating, and the collector current is low enough that the extrinsic resistances aren't important. ... Phenomenological models with a bunch of adjustable parameters can fit just about anything you like, and for more general circuit simulations they are certainly superior. ...
    (sci.electronics.design)
  • Re: OT: Bill Joy or Bill Gates?
    ... > comments on Moore's law is a good rebuttal to your point here. ... Moore's Law is about transistors and associated ... Predicting the future is iffy. ... Gates tossing a couple hundred million bucks for malaria nets and ...
    (rec.music.classical.guitar)
  • Re: The End of Moores Law
    ... "interpretation" of Moore's law is not correct. ... In his original paper, Gordon Moore claimed ... that "component" densities (not transistors, ... I was also informed that Moore's law has nothing to do with cost. ...
    (alt.lang.asm)
  • Moores Law never lived
    ... "interpretation" of Moore's law is not correct. ... In his original paper, Gordon Moore claimed ... that "component" densities (not transistors, ... I was also informed that Moore's law has nothing to do with cost. ...
    (comp.lang.asm.x86)