it is becoming obvious that A batchs are performing better than any of the few B's ive seen, including mine. No matter i do i cant get her past 3.48..
ad nasium....
Printable View
it is becoming obvious that A batchs are performing better than any of the few B's ive seen, including mine. No matter i do i cant get her past 3.48..
ad nasium....
I just got my Q6600 GO and I idle at 45 stock & 56 on orthos with a Tuniq tower with a Silverstone fan. Is this normal? How high is "safe".
I was looking at your fan 3 rpm on speedfan and was wondering wth is this rpm!!! turbocharger or what? :rofl:Quote:
Originally Posted by xgodxicex
hahaha i konw i dont get that....then again we all know how motherboard sensors can get...my favorite is my -5 and -12v voltagesQuote:
Originally Posted by Charloz24
I'm getting a difference of about 10C cooler from speedfan then core temp. Everest is also reading about 5C cooler then core temp. Temps are reaching 68-70CC under full prime95 load after about 2hrs under core temp. This is concerning me with such high temps. This is on a Q6600 G0 on AUTO V core. I'm not sure which temp to believe since they are all different. Any ideas? Oh yeah CPU is cooled using tuniq tower and 3 120mm, 1 200mm and 1 80mm case cooling. Idle temps are fine core temp reads about 35C idle such a big difference from idle to load with this quad over my old E6300.
http://img245.imageshack.us/img245/9...tledek4.th.jpg
First of all remove the AUTO setting for vcore and manually add a figure since AUTO usually adds more vcore than necessary.
What prime test are you performing, smallfft, blend etc? smallfft and largefft always stress the core more than blend test hence higher temps in coretemp.
Even though none of these programs can offer accurate temps trust coretemp.
Quote:
Originally Posted by raidflex
Actually, you're getting 15C cooler with SpeedFan, not 'about 10C' :). The short answer is that CoreTemp numbers are closer to being correct. The longer answer is that all of them might be wrong - we don't know for sure. Without rehashing this again, click on this link and start reading down through that page... for another few pages. It's all explained in bits and pieces there.
I did notice also that at about 70C the system will restart after about 3hrs of prime with no errors, I'm guessing its hitting the thermal threshold for a Q6600?
Possibly, but not necessarily... on CoreTemp 0.95, you should enable this option:Quote:
Originally Posted by raidflex
http://skinprops.com/images/ct095.jpg
...that will show you the value of the DTS, which counts down to zero as temps get higher. Zero is the throttling point of the cpu, so you can see how much 'temperature headroom' you have left. But as you overclock higher, you'll need more and more headroom to remain stable. My guess is that you're just not stable. Again, all of this (and more) is explained (mostly by uncleweb) in that link I provided above.
Thanks for the info I did read through that thread. So as it seems core temp is giving me the closest temp and that unless I can get my temps down that I may not get 3.6 stable.
Would it be worth it to lap the HS?
Only if it's not already flat. You could check it with a razor blade or other straight edge. You could also take another look at your case airflow - make sure all cables are tucked up out of the way, check your fans/speeds, etc.Quote:
Originally Posted by raidflex
Also, as suggested above, don't use Auto for cpu voltage... set it yourself, starting with about the reported VID (see CoreTemp), then increasing it until you get stable. The lower the voltage, the less heat you'll produce. You might also have to adjust other voltages (NB, etc). And check your memory stability, etc.
Case flow seems good I usually try to do the best job I can cleaning up the cables and tucking them in places where there are no fans for maximum flow. The HS I noticed does have some swirl machining marks but to me seems pretty smooth to the touch. From what I have been reading making a HS shiny has little affect on temperature as long as its smooth. Temps on the mobo are around 30C and if I run prime and then stop the stress test I noticed CPU temps drop right down to 40 and continue down to 35 almost instantly, this at least is telling me that the HS seems to be extracting the heat off the CPU effectively, but i could be wrong? One thing i also noticed is that between large FFT and small FFT in prime 25.3 there is no difference in temps or they are within a degree of each other.
My coretemp values are about 10 degrees celcius hotter than uGuru and 15 degrees higher than speedfan.
So I shouldn't trust uGuru?
My CPU is currently 100% stable at 3.2ghz with 1.395v in BIOS.
So my CPU is actually running close to 70degrees instead?
At 3.375ghz, my CPU goes over 80degrees in uGuru.
If coretemp is correct, that's about 90 degrees. Wow.
Now I don't know what to believe..... =(
Also, my FanEQ doesn't control my Tuniq. It did before. But now it doesn't.
ANyone have such issues? I haven't changed the cabling or anything. =(
http://img235.imageshack.us/img235/5...ablech2.th.jpg
Yeah that seems about right. My speed fan is about 15C cooler then what core temp and the Asus utilities say. Your board may be reporting it different then mine, i doubt its at 90C I would believe 80C although. If it was at 90C you would have instability problems probably.
I run mine at 3.4G sealed in the case,it takes 1.537 v core and idles at around 50c and 30 min occt load at hottest is 87c in core temp 0.95
PWM peaks at 79c in uguru with two 120 cm yate loons glued in the case at the cpu/pwm end blowing across the board, do you Think it will run like this for 12 to 18 months at least
Q6600 is a g0 L723A930 with a VID of 1.1625 ordered from ncixus on 7-25
At 87C load with 50C idle i don't see it lasting that long. It really does appear that air cooling needs to step up a bit to handle these quads.Quote:
Originally Posted by TV Addict#2
6 months would be enough
seeing as it doesn't throttle until 100c it may handle the heat pretty well,the test is in progress
documenting total run time,course few game will stress it to that extreme though
anything above 75C is high. Although this is under extreme load of prime which is different no programs that i know of will stress the CPU that much. The idle temps do seem high though, mine idles about 35C maybe you should try reseating the CPU. At those temps you would be lucky to get 6 months out of it or most likely it will slow degrade over time and you may get errors and instability, I would try and keep it below 70C on regular use and you should be fine.
I ordered one from NCIXUS and it is a G0. When it gets here, i'll show the results.
Well, lapped my Q6600 B3... running 450x8 1.55v not to shabby load temps are 57-58C priming now.
I got it under control lapping my Ultra-120Extreme, and now if it passes 24hr I will lap the chip.
Also have a G0 on order.
EDIT: Anyone else find Prime95 25.3 to be really touchy? Its very picky on what is stable and what isnt.
Orthos with Core Affinity set to SmallFFT runs forever..... wonder if I should trust Prime95...
i just installed a thermalright ultra 120 extreme with shin-etsu paste.
my ninja was a POS compared to this cooler.
my load temps with the ninja were hitting 69/71 @ 3.4ghz with the same volts as my loads now @ 3.42ghz being 10c less at 61/59. ambient temps have been more or less the same, high 70F in the room, and humid.
i also idle 5 degrees lower.
i benched this at 3.5 for 15 minutes which is nothing, and at 3.42 first for an hour - all stable of course - but i gotta goto bed its late eastern time and worky in the mornin.
i expect to be stable close to 3.6 or higher but ill run this 24/7 at 3.42 (im no xtreme oc'er) since my idle is nice and 61 load is dandy with me since real world usage will never get there - and provide an updated OC after hours of prime stable on my high oc.
with both coolers i ran an antec tri-cool on high with the rear exhaust on high (due to summer)
in fall/winter i run the fans on medium or low.
Quote:
Originally Posted by ChaosMinionX
With my Q6600 B3 @ 3.6Ghz with 1.42v real was Prime95 25.3 stable using small ffts but would fail large ffts fauirly quickly. Bumped the vcore to 1.44v real and large ffts passed for 6 hours.
Also ran Orthos with affinity set to run all 4 cores and used Large ffts - Priority 9 as the ultimate stability test and it passes that fine so in my eyes it's stable as it ever will be.
This B3 must just really suck haha....Quote:
Originally Posted by lowdog
Hell I cant even get 3.2ghz stable on my B3....BOOOOOOO:(