Did I do something wrong? E6600 OC Temp

[soulesschild]

My E6600 is currently OCed to 3262 with Vcore at 1.42V, the load temp under an orthos small FFTs test is 70 C. I have a Zalman 9700 on it. Is this normal or did I not apply the fan/heatsink/thermal paste correctly?

 

[soulesschild]

okay after some thinking, i redid the thermal paste on my CPU and the temps dropped quite a bit .

Orthos small FFTs test, 54 C from 70-73 C

idle @ 33-40 (dont know why theres a huge gap), from 45-55 C


Guess everybody has dumb mistakes ><....now if i could only clock 3.4ghz stable....

 

[davidlem]

So...what did you change??

 

[hmz]

I'm getting weird temps too with stock HSF. Idle 53C which is too high, and load in Orthos ~68C. I'm sure the HSF is mounted correct because i inspected the back of the MB. I'll get Zalman soon.

 

[jmackay]

Zalmans aren't that great. The best HS for the C2Ds are Thermalright, Tuniq, or Arctic Freezer Pro 7.

 

[Imitation]

I've seen a couple builds from my friends where they didn't get one of the 4 posts all the way in. Both were running really high idle temps and loads jumped real quick. Both double checked the pins and found one that wasn't all the way down, fixing them fixed the temps. Both were stock coolers.

 

[hmz]

I've seen a couple builds from my friends where they didn't get one of the 4 posts all the way in. Both were running really high idle temps and loads jumped real quick. Both double checked the pins and found one that wasn't all the way down, fixing them fixed the temps. Both were stock coolers.

Ya, but its too much work to take out MB from the case, disconnect all the cables. I would do this when i get new HS.

 

[soulesschild]

while the zalman may not be the best, it does its job better then stock

 

[AliP]

I have a very similar setup, but my load temps only get to a max of 50c, (also 9700+680i+e6600@3,2)

 

[arentol]

Zalmans aren't that great. The best HS for the C2Ds are Thermalright, Tuniq, or Arctic Freezer Pro 7.

I ran a statistical analysis of the results listed in the C2D overclocking thread as of 6 weeks ago, and based on those results I would have to say you are wrong.

I analyzed all HS's with at least 3 results in each category, e6300, e6400, and e6600. I looked primarily at increase from stock speed for each processor.

Here are the results...

Cooler --- Top Average Increase --- Highest Increase --- Average Increase
Turniq 120 --- 1491 --- 1871 --- 1426
Zalman 9700 --- 1357 --- 1728 --- 1322
Zalman 9500 --- 1344 --- 1605 --- 1124
Thermalright Big Typhoon --- 1342 --- 1648 --- 1141
Scythe Infiinty --- 1296 --- 1538 --- 1232
Scythe Ninja --- 1292 --- 1570 --- 1082
Arctic Freezer Pro 7 --- 1222 --- 1640 --- 1054
Stock --- 1137 --- 1451 --- 964

Top Average Increase = Top 3 results in each category combined and divided by 9.
Highest Increase = Highest Increase from stock among all categories
Average Increase = Average Increase among all results

I rank them according to the Top Average Increase because generally speaking the top 3 are all ones that were pushed to their limit whereas the Average includes many that were clearly taken only to a nice round number, like 3200, and then stopped even though they could have gone further. (Interestingly enough as of the time I ran this all three of the highest results for the Turniq and the 9700 e6600's were 3600MHz.. Indicating nobody had yet tried to push either of these cooling system to their limits with that CPU and yet they still beat all other comers overall).

Regardless, as you can see it is quite clear that the Turniq complete dominates, with the Zalman 9700 a distant 2nd, the Zalman 9500 and Thermalright tied for 3rd, the Scythes tied for 5th, and the Arctic Freezer being a distant 7th place out of 8 finishers!

 

[alik4041]

Zalmans aren't that great. The best HS for the C2Ds are Thermalright, Tuniq, or Arctic Freezer Pro 7.

Agreed. I honestly think Thermalright products are underestimated. BTW, anyone know when the Thermalright Ultra 120 Extreme will be available? From this review, it seems amazing:
http://www.anandtech.com/casecooling/showdoc.aspx?i=2943

We asked at the beginning of this review if more is better. In the case of the Thermalright Ultra 120 Extreme the answer is a solid "Yes it is". The Tuniq Tower 120 and original Thermalright Ultra 120 remain very competitive heatpipe towers. In that same ball park are the Cooler Master Hyper 6+ and the Scythe Infinity with push-pull fans. You will not be disappointed in any of these coolers. However, the best air cooler we have tested is the Thermalright Ultra 120 Extreme. It cools better at all speeds and overclocks, and overclocks the processor further than any air CPU cooler we have tested. We look forward to seeing it on retail shelves, and performance enthusiasts should definitely try to find an Ultra 120 Extreme to cool their processor when it hits the retail channel.

 

[jmackay]

lots of #s

What do those numbers even mean?

Increase in what? Temp? Speed?

And here are what some of the experts think:

http://www.anandtech.com/casecooling/showdoc.aspx?i=2943&p=4
http://www.silentpcreview.com/article727-page3.html

To quote SPC in their Zalman 9700 article:

When it comes to performance in the context of low noise, the 9700 doesn't fare well against the competition of the big towers and the reference quiet 120mm fan. Add the higher price tag to the mix, and the match is over.

So dunno what your numbers mean, but I know what Celsius is and I know that LESS IS BETTER.

 

[soulesschild]

i dont notice the noise that much and my zalman 9700 NT was about 35 dollars from a mom and pop store i bought it from (some special coupon) so i'd say i did okay

 

[hmz]

Tuniq is huge and Arctic Freezer Pro 7 it is pain to install, so i choose Zalman.

 

[jmackay]

Tuniq is huge and Arctic Freezer Pro 7 it is pain to install, so i choose Zalman.

You are kidding right?

The Arctic Freezer Pro 7 uses the standard mounting just like stock. To mount a zalman you need to remove the mobo to install a backplate just like the Tuniq/Ultra-120 and most other heatsinks.

 

[arentol]

What do those numbers even mean?

Increase in what? Temp? Speed?

And here are what some of the experts think:

http://www.anandtech.com/casecooling/showdoc.aspx?i=2943&p=4
http://www.silentpcreview.com/article727-page3.html

To quote SPC in their Zalman 9700 article:


So dunno what your numbers mean, but I know what Celsius is and I know that LESS IS BETTER.

Those numbers were based on the results printed in the thread below as of about 6 weeks ago...

http://www.hardforum.com/showthread.php?t=1138241

I will start with the SECOND number listed after each HSF, or the "Highest Increase". You will note that the Highest Increase listed for the Tuniq 120 is 1871. If you go to the thread above and look at the second result in the e6300 section you will see that it is an overclock by Donkers of an e6300 to 3731MHz This is the SINGLE highest overclock above default among all Tuniq 120 overclocks reported. If you subtract the default clockspeed of an e6300, 1860MHz from 3731 then the result is 1871 MHz. So that is the HIGHEST INCREASE FROM STOCK AMONG ALL CATEGORIES TESTED (e6600, e6400, and e6300) for that HSF... Same for all other HSF's.

The FIRST number after the name of each HSF was found by doing the same process described above, but doing it for the 3 highest overclocks listed for each HSF in the three categories, e6600, e6400, and e6300. Those totals from EACH category were added together and divided by 9 (3 results per category times 3 categories = 9 results). So the First number is the AVERAGE of the top 3 results from each of the 3 categories tested.

The THIRD number after the name of each HSF is just like FIRST number, except that is the average of ALL results for that HSF among all 3 categories and all reports given.

However, all that is kind of irrelevant. The point is that if you do an actual analysis of the numbers available it is quite clear that the Zalman 9700 is at least as good as anything out there except the Tuniq. Yes, the tests and the results are not completely scientifically accurate, but the end result is that all these top coolers are competitive, and claiming the Zalman's are not so is just plain wrong.

 

[alik4041]

okay after some thinking, i redid the thermal paste on my CPU and the temps dropped quite a bit .

Orthos small FFTs test, 54 C from 70-73 C

idle @ 33-40 (dont know why theres a huge gap), from 45-55 C


Guess everybody has dumb mistakes ><....now if i could only clock 3.4ghz stable....

Ya... I was wondering how 70C was possibel at 1.43v w/ a Zal 9700 and an anty 900 lol. Good job fixing your problem. I'm guessing it was not enough thermal paste? BTW, what compound you using? AS5?

 

[arentol]

What do those numbers even mean?

Increase in what? Temp? Speed?

And here are what some of the experts think:

http://www.anandtech.com/casecooling/showdoc.aspx?i=2943&p=4
http://www.silentpcreview.com/article727-page3.html

To quote SPC in their Zalman 9700 article:


So dunno what your numbers mean, but I know what Celsius is and I know that LESS IS BETTER.

As for those reports... the AnandTech one is all over the place with OC results relative to measured temperatures.. Kinda makes you wonder WTF happened there, because in my experience cooler = higher OC, and yet that was not consistently the case for them.

Regardless, the actual temps on the 9700 in that test are the 4th best behind the Thermalright Ultra 120 and 120 Plus, and the Tuniq. This indicates to me it is highly competitive no matter how you look at it.

As for the Silent PC review...That is about SILENCE, not actual PERFORMANCE when overclocking so I am not quite sure why you think that is relevant.

Also keep in mind I am NOT recommending the Zalman. I strongly believe in the Tuniq, and the Thermalright 120 Plus and don't see why anyone would bother with anything lesser.

However, I do think a blanket statement that ALL Thermalrights are better than Zalmans is clearly wrong, and I don't see how the Arctic Freezer Pro 7 even got into the discussion.

 

[AndyM]

On the OC thread...really depends on if someone wanted to post a 75C overclock or not . FWIW, the Tuniq is a good cooler. But a better test of 'which cooler is best' would be made if all the reviews would lap the IHS and cooler flat, introducing varying mating surface variables in published tests kind of defeats the purpose of doing the test. As an example my IHS and Tuniq were nowhere near flat resulting in the cooler being lifted off the center of the IHS where the die resides below. If this was a published test...the Tuniq would suck. But lapping both dropped load temps 10C, so if that could be compared to similarly prepped coolers then there would be a fair comparison. Of course making sure fan speeds and sizes were appropriately matched or if that could not be done, dB readings matched up so that a fair comparison could be made, would be quite beneficial in making a cooler choice based on the reviews. My .02

 

[idiotekniQues]

I ran a statistical analysis of the results listed in the C2D overclocking thread as of 6 weeks ago, and based on those results I would have to say you are wrong.

I analyzed all HS's with at least 3 results in each category, e6300, e6400, and e6600. I looked primarily at increase from stock speed for each processor.

Here are the results...

Cooler --- Top Average Increase --- Highest Increase --- Average Increase
Turniq 120 --- 1491 --- 1871 --- 1426
Zalman 9700 --- 1357 --- 1728 --- 1322
Zalman 9500 --- 1344 --- 1605 --- 1124
Thermalright Big Typhoon --- 1342 --- 1648 --- 1141
Scythe Infiinty --- 1296 --- 1538 --- 1232
Scythe Ninja --- 1292 --- 1570 --- 1082
Arctic Freezer Pro 7 --- 1222 --- 1640 --- 1054
Stock --- 1137 --- 1451 --- 964

Top Average Increase = Top 3 results in each category combined and divided by 9.
Highest Increase = Highest Increase from stock among all categories
Average Increase = Average Increase among all results

I rank them according to the Top Average Increase because generally speaking the top 3 are all ones that were pushed to their limit whereas the Average includes many that were clearly taken only to a nice round number, like 3200, and then stopped even though they could have gone further. (Interestingly enough as of the time I ran this all three of the highest results for the Turniq and the 9700 e6600's were 3600MHz.. Indicating nobody had yet tried to push either of these cooling system to their limits with that CPU and yet they still beat all other comers overall).

Regardless, as you can see it is quite clear that the Turniq complete dominates, with the Zalman 9700 a distant 2nd, the Zalman 9500 and Thermalright tied for 3rd, the Scythes tied for 5th, and the Arctic Freezer being a distant 7th place out of 8 finishers!

you did all that work and it doesnt mean squat.

you arent taking into account the variables of chip week, stepping etc plus all the variables in the user, like how far did they really push the chip, what kind of stability did they test for. that makes judging the coolers by this data worthless practically.

your entire analysis is useless.

 

[jmackay]

As for those reports... the AnandTech one is all over the place with OC results relative to measured temperatures.. Kinda makes you wonder WTF happened there, because in my experience cooler = higher OC, and yet that was not consistently the case for them.

Regardless, the actual temps on the 9700 in that test are the 4th best behind the Thermalright Ultra 120 and 120 Plus, and the Tuniq. This indicates to me it is highly competitive no matter how you look at it.

As for the Silent PC review...That is about SILENCE, not actual PERFORMANCE when overclocking so I am not quite sure why you think that is relevant.

RTFA

Using a SLOWER (quieter) fan than the Zalman, it STILL beat the zalman by 4-5c. Considering you can stick a monster of a fan on the ultra-120 if you don't care about noise, it would beat the crap out of it.

Taking overclocks from different people doesn't mean squat. Places that review coolers use the same setup, same room temp, same case fans, etc.

Joe with 8 250 fans blowing on his Zalman to make it cooler than Bobs Tuniq/Ultra-120 doesn't mean the Zalman is better.

 

[snaggletooth]

my e4300 at 3.0ghz idles at 27C as reported in speedfan 4.32 and is mirrored in PCProbe2

ACF7 with AS5


that's a ton of work for all that HSF testing! please forgive me if i find your ACF7 results hard to believe.