[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: server hardware
On Sun, 15 Nov 1998 00:09:49 +0200, Nahum Greenberg <green@pf1.co.il> wrote:
>tuvia beker wrote:
><...>
>> Note thatthat the 300A celerons are not only the most overclockable chips
>> ever made - the fact that their cache is on-dye compensates for its
>> reduced amount, and this compensation is even more noticeable at 450MHz.
>> In fact they are only slightly weaker than regular P450.
>> I seriously consider doing it, mostly for the fun of it...
><...>
>Yeah, sounds like fun indeed. Me and a couple friends of mine are going
>to SMP-ize the Celerons in early December. However, in my opinion it's
>better to use regular Celerons, not the A-type: they're cheaper, you can
>get a used ones and the multiplier's not locked, like on the newer Intel
>chips.
WRONG.
Easrly Celeron 300's had selectable multiplier, but according to tom's
hardware page, this was blocks a few weeks after he published this information
in the first place.
Me and a friend of mine have overclocked a 300A from 300MHz to 450MHz, but in
order to get it to work stable, we had to increase core voltage from 2.0V to
2.05V. This required making a tiny short between 2 pins on the celeron board.
We are still not sure it's 100% stable, even though it works for hours,
running heavy load disk/video benchmarks (about 20 copies running in
parallel). We MAY raise it to 2.1V if necessary.
If you have an Abit 440BX motherboard with BIOS selectable core voltage, you
don't need this. (On the other hand, a malicious Virus can fry your CPU...)
Also, the CPU got quite hot (untouchable) with the standard single fan. We
ended up adding an extra 12V fan from an old power supply, which really cooled
it. We installed the fan on the BACK of the CPU.
Also, try getting the BOXED versions of the CPU. They cost the same as an OEM
Celeron + Taiwanese fan (maybe 1-2 dollars more), but the Intel heatsink+FAN
is MUCH BETTER than the typical junk you get. Don't forget that power goes up
linearly with frequency, so you expect a 50% power increase.
>The real question is whether or not the penalty for absence of the cache
>will be really that great? Obviously, it will depend on application. For
IMHO, it's worse enough on single processor systems. With SMP, it will be MUCH
WORSE. Don't forget that with a dual processor board, the memory is a shared
resource for BOTH CPU's, so any cache miss is much more expensive, and the
memory bus is saturated much easier, even with 100MHz SDRAMs (800MByte/s).
As for myself, I'm getting a 300A Boxed Celeron and a new Tekram motherboard
tomorrow. I know it's not fast (see toms hardware), but I get it for $98, and
it has a management chip for monitoring core voltage, and fan speed. An ASUS
board costs about $170.