my 30 capper 2.8c SL6Z5 runs alot warmer than my 2.4c SL6Z3-WHY???

[Jojo1971]

AT FULL LOAD, my 30 capper 2.8c SL6Z5 (64°C-68°C at full load). runs about 14°C-18°C warmer than my 2.4c SL6Z3 (max of 50°C at full load).

using the same mobo, HSF and vcore..room temp is controlled...also tried in another rig (XPC)...tried reseating the processor too..the results are the same...WHY??? What could be the reason?

does my 2.8c MO have "prescott blood" in it???

[Lancelot]

Read my signature, I get 50C full load with 21C ambient here right now. Are you using the same TIM as with your 2.4C?

[Nebulous]

Umm....... have you let the TIM cure? After switching cpu's, cleaning off the used TIM, cleaning up the hsf, applying fresh TIM, then reseating the hsf will 99.99.99% of the time result in higher temps until the TIM cures and settles. After switching my 2.6c @ 3.6GHz 1.6v loaded @ 38c to the 3.0E i have now, running 3.75GHz with 1.45v loaded temps are 41c. This is after 48 hours of Folding/HotCpu/Toast. temps b4 were 48c loaded on the pressy.

Did you allow enough time for this process?

[Jojo1971]

Umm....... have you let the TIM cure? After switching cpu's, cleaning off the used TIM, cleaning up the hsf, applying fresh TIM, then reseating the hsf will 99.99.99% of the time result in higher temps until the TIM cures and settles. After switching my 2.6c @ 3.6GHz 1.6v loaded @ 38c to the 3.0E i have now, running 3.75GHz with 1.45v loaded temps are 41c. This is after 48 hours of Folding/HotCpu/Toast. temps b4 were 48c loaded on the pressy.

Did you allow enough time for this process?

my findings are from more than 3 weeks of observation....

[Jojo1971]

Read my signature, I get 50C full load with 21C ambient here right now. Are you using the same TIM as with your 2.4C?

yes im using the same tube of AS5...

[hUMANbEATbOX]

not all cpus are created equal.

if the 2.8 is a newer one, they are rumored to not be as good as oler m0's, but i dunno if there is any truth to that.

[larva]

This is another one of those topics that is almost uniformly misunderstood. I have posted this information many times, with people understanding its accuracy and importance as the exception rather than the rule.

Heat numbers are slighty useful in a comparitive sense, but absolutely useless in an absolute one. CPUs do indeed vary, and in every way including the temperatures read. The cpu's core may actually be no hotter than another chip that reads 5c cooler, or maybe it is. Who cares. The real point is how much clock speed the chip proves stable at. Better chips will be stable at high clock rates in spite of medium to high temps, if the cooling makes or breaks the OC it often wasn't meant to be.

My 30cap 2.8c reads what I consider to be on the warm side. But I did not buy it to produce a 45C loaded temp, I bought it because it was proven stable at 3.5GHz and that is good enough for me considering it allows me what I consider an optimal fsb/memory/timings condition. The chip has potential at 3.6GHz, but clearly needs more cooling to do this. I'm just not concerned with pressing the chip that hard (and listening to the noise of the fan that cools it).

My previous chip, a C1 1.8a, needed 1.75V to run at even 3.33GHz. I was expecting my 2.8 (which needs only 1.57V at 3.5GHz) to run cooler. It didn't. O well. The true indicator of the worth of a chip is it's clock speed potential, read temperature is more of a curiosity.

Overclockers, having a true do-it-yourselfer attitude, wish to believe they can transform the results by dumping money and effort in the top. And to an extent they can. One primary tactic towards this end is drastic cooling (and the hopes of really low numbers in MBM) in the uniform expectation of drastic clock speed rewards.

The simple fact is the biggest factor is how good of a chip you have. You could very well scrabble about with an average chip and awesome cooling at 3.6GHz where a better chip might run great at 3.7 with the retail cooler reading 65C in the bios... I maintain that more progress is made through isolatiing the really good chips than twisting the arm of poor ones.

Now in the case of the P4 I must qualify these remarks by mentioning the throttling aspect of the design. Around 75C they will throttle, but I'm not certain they know or care what the bios reads. They number may be in error and the throttling mechanism working on better information, for all I know. In any event you need to verify that you are not throttling under prolonged 100% cpu load. If you aren't and the chip is stable, the temperature number is not an operative parameter.

It wasn't all that long ago that we had no temperature monitoring whatsoever, and to be honest we didn't want for it.

PS-Let me add that obviously cooling is an important aspect to system design and optmization. But just cranking up the cooling doesn't make the difference in all situations, and at times we are chasing a pink elephant anyway due to bad information. One should always strive for the most effective cooling possible within the envelope of allowable noise for both performance and durability reasons. But one certainly can't count on drastic cooling to transform all situations, and it isn't nesecessary in all cases.

[Jojo1971]

i guess i got a really good 2.8c, considering it is rock stable at 3.6 with default vcore and medium-to-high temp @ full load.... :)

[pvnarc]

where did you get your 2.8c's, i cant find anywhere that sells them with m0 stepping

[Starter13]

Put a SP120 then you sould get much less :clap:

[Zebbo]

Put a SP120 then you sould get much less :clap:

Actually, it's not SP120, it's XP-120 ;)

[GigaHertzAddict]

AT FULL LOAD, my 30 capper 2.8c SL6Z5 (64°C-68°C at full load). runs about 14°C-18°C warmer than my 2.4c SL6Z3 (max of 50°C at full load).

using the same mobo, HSF and vcore..room temp is controlled...also tried in another rig (XPC)...tried reseating the processor too..the results are the same...WHY??? What could be the reason?

does my 2.8c MO have "prescott blood" in it???

This is normal. Different CPU's, different temps. the 30 cap CPU's run crazy hot compared to older ones.

[Jojo1971]

where did you get your 2.8c's, i cant find anywhere that sells them with m0 stepping

fortunately saw an ad in AT forums for a retail 2.8c for $125..when i asked the seller for the codes i was happy when he told me SL6Z5..of course i sent my payment in a heartbeat ...pack date is 5/04.. i was skeptic at first about it's OCing ability coz ive always thought that the older MO's are the better ones..well, when i fired it up, i was surprised...

im starting to gather watercooling parts again..i hope i can get this stable at 3.8GHz with WC...

[Private Joker]

Btw. What happened to your great C1 1.8a, larva?
I allways looked with astonishment at your Sig.. :drool:

[larva]

I've still got it, it's in my tv-tuner pc now at 2.7GHz. It lived a hard life in my primary rig, but has it easy now. I ran that thing at 1.75+V for a year and a half...lol. You'll pardon me if I don't live in fear of SNDS.