Re: LZ: optimizing zshell programs for


[Prev][Next][Index][Thread]

Re: LZ: optimizing zshell programs for



On Thu, 5 Sep 1996, SAM STAUFFER wrote:


> I agree with that, but for example you could speed up the Pentium 70Mhz
> to 90 or even 120 by just raising the clock speed. They make all
> 70,90,120 to be the same speed, but they test them at 120 until they
> meet their quota, then 90, and then the rest are 70s. Some 90s don't get
> tested as 120, and still might work. and most 70s are never tested at
> all, so they could be 90s or 120s.
> 
> It doesn't wear out the processor even if a test fails at a higher
> speed, if it did they couldn't keep testing them at lower speeds.
 
>From what I've read, the higher the speed the higher the
temperature.  The ones that pass at 120 don't get as hot
as the ones that fail.  Running hotter does shorten the
life of the chip but a certain amount of reduced life is
thought acceptable for the higher speed.


The problem is that if you have one of those that didn't
pass at higher speeds and you increase it's speed then
you're shortening it's life more than the Intel engineers
think is acceptable.  You might not care, though.  I have
no idea what kind of reduced life is involved, but it's
enough that Intel doesn't want it's reputation to be
affected by it.


I also have no idea whether the same priciples apply in
the Z80 in these calcs, although I suspect they do.  It
might be that these chips are rated conservatively.  Or
not.  Who knows?  You pays your money and you takes your
chances.


Barry


References: