Blade Servers

Lennart Sorensen lsorense-1wCw9BSqJbv44Nm34jS7GywD8/FfD2ys at public.gmane.org
Wed Feb 16 20:28:59 UTC 2005


On Tue, Feb 15, 2005 at 06:12:38PM -0500, Byron L. Sonne wrote:
> I'd be careful about the use of the word 'simply'... I don't design 
> cores or fab chips for a living, but I'm gonna have to guess that it's 
> probably fiendishly complex. Intel doesn't fuck things up because 
> they're stupid or lazy ;)

Pentium fdiv bug.  I think that one might qualify as being lazy.
perhaps a complete verification of the table used to speed up fdiv would
have been in order.  Not doing so to me sounds almost like someone was
lazy.

As for stupid, I honestly believe intel let their marketing department
dictate the design goals of the P4, which runs high clock rates, while
not getting as much done per clock as other similar generation chips,
while using lots of power and making a ton of heat.  The Pentium M at
lower clock rates runs as fast or faster on many applications, using
less power, making less heat, and it based on an older design.  Too me
that makes whoever thought just aiming for winning the clock speed game
was indeed stupid.  And they must have worked for intel at the time.

Well that's just my opinion on the matter at least.  intel isn't going
out of business, and they do know how to design good chips.  They aren't
all good though, and sometimes they do miss what the market wants (or
doesn't want, as may be the case for the itanium.  I am not really sure
what is wrong with the itanium, besides the price and lack of software).

Lennart Sorensen
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml





More information about the Legacy mailing list