Computer Chess Club Archives




Subject: Re: Introducing "No-Moore's Law"

Author: Steve J

Date: 17:45:54 02/27/03

Go up one level in this thread

On February 27, 2003 at 19:16:41, Jeremiah Penery wrote:

>On February 27, 2003 at 18:09:47, Robert Hyatt wrote:
>>On February 27, 2003 at 16:51:44, Jeremiah Penery wrote:
>>>On February 27, 2003 at 13:31:54, Robert Hyatt wrote:
>>>>What you have written can't be true.  Because if you read this forum long
>>>>enough, you realize that many are convinced that designing chips is a trial->>>and-error process where the engineers don't know anything at all about how >>>fast a chip will run until it is produced and tested.  No ideas about the >>>expected wafer defect rate.  Etc.
>>>Nobody ever said or implied anything of the kind.  You just seem to think that
>>>CPUs are produced as near that theoretical clock limit as possible, which is
>>>simply not true.
>>Sorry, but it _is_ true.  Yes, parts are marked "down" when there is a demand,
>>but ask an engineer that works for one of these companies.  They don't design
>>something and hope it will run fast.  They know _before they start_ how fast >it is going to clock within a small error margin.
>I hope you're not taking reading comprehension lessons from Vincent or anything.
> One of him is enough.  I'm not disputing that they know how fast it should
>theoretically clock.  I am disputing the assertion that they actually sell them
>within even 10% of the limit.  Usually, they don't.
>>>>It was a nice explanation of a known issue, but it can't be right because it >>>implies that the engineers really know what they are doing, rather than >>>relying on blind luck to get something to work.
>>>>Of course, all the engineers I personally know are repeating your story and >>>they are sticking to it.  But they must all be mistaken.  After all >>>engineering isn't a true science, it is based mostly on serendipity.
>>>Not that any of what you say here has anything to do with Moore's Law in the
>>>first place...
>>Indirectly it does.  Moore's law relates to the size of a transistor.  But that
>>is also proportional to how fast the thing will clock.  If it goes faster, it >will be smaller. If it is made smaller, it will go faster.  Circular, IE...
>All you said in the previous post was that engineers know how to make chips.
>Not a word about shrinking transistors.
>"If it goes faster, it will be smaller."  Where does that come from?  Making it
>smaller allows it to go faster, but making it faster does not imply that it is
>made smaller.
>>>Engineers can't predict the future.  They may be great at telling us the
>>>limitations of current technology, but they can't guess about the emergence of
>>>new technologies.  Many engineers thought the sound barrier was impossible to
>>>break, but that didn't make it true.  They've been saying for years that >>Moore's Law will fail.  Here we are today, with no indication of slowdown in >>the next few years.
>>I disagree with that last.  If you look at the clock speeds, things are _not_
>>progressing as fast now as they were 5+ or 10 years ago.  For the reasons
>>given already.  Pathways are _not_ going to become sub-atomic in their
>>width.  So there is a clear asymptote on the horizon...  How close to a
>>single atom they can go might be open for argument, but not beyond that.
>They're not all that close to even the molecular limit, let alone the atomic
>one.  There's promising research into optical relays that can detect streams of
>single photons sent at high frequency.  Eventually, something like this could
>replace the current electronic pathways (wires) in a CPU.  Anyway, the point is
>that you can't predict what future technology might accomplish.  History is
>littered with the remains of those who built their world view on the assumption
>that something was impossible, only to see the 'impossible' become everyday
>>>There are many possible ways its usefulness can be extended past what many >>currently believe is possible.  Examples are finding a new manufacturing >>process that allows much smaller features to be created, finding a better >>material than silicon, etc.  It's possible that a completely new computing >>paradigm may become usable, rendering Moore's Law completely obsolete.
>>>Examples of this may be DNA and/or Quantum computing.

Robert and Jeremiah,

  Thanks for the posts.

  One point I was trying to make was that every reduction in size is done
with an exponential increase in cost.  We will reach a time when the
physics of very small devices will not allow for transistor that can be
turned "on" and "off" at any reasonable cost.  Given that this is related
to the size of the atom, it does not make much of a difference if the
material is Silicon, GaAs, InP, or more exotic materials.

  As Robert had mentioned, increasing the number of processors will then
be the most effective way of increasing NPS.

  Of course, this applies to the semiconductors and hardware only.  Software
improvements, however, will continue exponentially for all time.  :)


>>You can't make a feature smaller than the molecule you are working with, if the
>>substance is molecular.  Can't go below the size of an atom with any forseeable
>>process.  Whether quantum computing comes to pass or not is another subject,
>>of course.  But for _current_ computing, there's a definite limit on the >horizon with no forseeable workaround.
>For current technology and method of CPU manufacture, yes.  I have every
>confidence that some way will be found to circumvent the limitations, as has
>been done innumerable times in the past.  At this point, all we can do is guess.

This page took 0.06 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.