Yes. You are right. This is what I have been setting. 2500MB is the free ram and not the available.
Edit: Are you using one fan in the middle or have you placed an extra fan in front of the heatsink?
Last edited:
Pea sized blob method for Q6600?
- Thread starter WingZero30
- Start date
Yes. You are right. This is what I have been setting. 2500MB is the free ram and not the available.
Edit: Are you using one fan in the middle or have you placed an extra fan in front of the heatsink?
OK guys here are my final IBT results for [email protected]. After this I don't want to kill my cpu 
.
Threads was set to 4
GFlops value increased slightly to over 44 GFlops roughly which is 80% of the theoretical maximum of 3.4 x 16 = 54.4GFlops. So I am happily satisfied with the results
.
Temps increased slightly by 2-3C due to increase in GFlops.
Again the test took 1 hour 26 mins roughly.
So now I have experiemented with different thermal paste application methods (MX-4) and ran Intel Burn Test several times ensuring I obtained higher GFlops values, so sort of exhausted the cpu stress testing experimentation. The only thing left is Prime Blend test for overall stabilty.
I could go up to 3.5Ghz with Prime95 but not with IBT and likewise for 3.6Ghz as coretemps will definitely reach over 80c. The only way I think I may be able to achieve 3.6Ghz is if I use Asus P5Q Deluxe mobo and custom water cooling. But I think for extra 200MHz I don't think it's worth spending the money as I won't notice any difference between 3.4Ghz and 3.6Ghz in gaming which I usually do on pc, unless I am benching.
.Threads was set to 4
GFlops value increased slightly to over 44 GFlops roughly which is 80% of the theoretical maximum of 3.4 x 16 = 54.4GFlops. So I am happily satisfied with the results
.Temps increased slightly by 2-3C due to increase in GFlops.
Again the test took 1 hour 26 mins roughly.
So now I have experiemented with different thermal paste application methods (MX-4) and ran Intel Burn Test several times ensuring I obtained higher GFlops values, so sort of exhausted the cpu stress testing experimentation. The only thing left is Prime Blend test for overall stabilty
.I could go up to 3.5Ghz with Prime95 but not with IBT and likewise for 3.6Ghz as coretemps will definitely reach over 80c. The only way I think I may be able to achieve 3.6Ghz is if I use Asus P5Q Deluxe mobo and custom water cooling. But I think for extra 200MHz I don't think it's worth spending the money as I won't notice any difference between 3.4Ghz and 3.6Ghz in gaming which I usually do on pc, unless I am benching.
You worry too much mines hit 85-95 c on IBT and gaming temps still under 60c
3.6ghz is surely quite a bit faster for gaming no ? stock fan in the middle on tuniq I have experimented in the past by adding a fan at each side and it doesnt improve temps at all strangely.
the tuniq is capable of running it 3.6ghz for gaming and prime but just not for IBT loads imo prime will go 60-65c even at 1.45v but IBT just creeps up to throttling temps of 80-95c
I reckon with that cooler you can get 3.6ghz for 24/7 easily IBT as long as 80c peak is fine
People have ran q6600's 24/7 folding with 1.5 volts for years
3.6ghz is surely quite a bit faster for gaming no ? stock fan in the middle on tuniq I have experimented in the past by adding a fan at each side and it doesnt improve temps at all strangely.
the tuniq is capable of running it 3.6ghz for gaming and prime but just not for IBT loads imo prime will go 60-65c even at 1.45v but IBT just creeps up to throttling temps of 80-95c
I reckon with that cooler you can get 3.6ghz for 24/7 easily IBT as long as 80c peak is fine
People have ran q6600's 24/7 folding with 1.5 volts for years
Last edited:
I think one of the reason that IBT is more stressful is because it's a cyclic loading program in nature. Initially when it starts there is moderate loading on the cpu and then suddenly the loading skyrockets and temps increase tremendously.
When it outputs result for each run, the loading is momentarily decreased even though coretemp is still reporting 100% cpu load and temps become lower and then the cycle repeats. This imo puts cpu under greater stress and you can think of it as being akin to a light aluminium or a light metallic road which you constantly turn and twist until physical crack appears.
Prime95 is more useful for long term stability as the loading is constant at 100% though not as intensive and temps are as a result modestly higher. With IBT you don't know whether your overclocking stability crashed due to excessive heating or lower vcore.
When it outputs result for each run, the loading is momentarily decreased even though coretemp is still reporting 100% cpu load and temps become lower and then the cycle repeats. This imo puts cpu under greater stress and you can think of it as being akin to a light aluminium or a light metallic road which you constantly turn and twist until physical crack appears.
Prime95 is more useful for long term stability as the loading is constant at 100% though not as intensive and temps are as a result modestly higher. With IBT you don't know whether your overclocking stability crashed due to excessive heating or lower vcore.
Last edited:
Hi guys
What settings should I choose for prime blend test? Shall I just choose the default settings i.e 8k to 4096k and 1600MB ram or should I alter these for greater stress testing? I have 4GB ram (2x2GB) in dual channel mode and 'free' physical ram is always usually around 2700MB in windows task manager when windows has loaded.
many thanks.
What settings should I choose for prime blend test? Shall I just choose the default settings i.e 8k to 4096k and 1600MB ram or should I alter these for greater stress testing? I have 4GB ram (2x2GB) in dual channel mode and 'free' physical ram is always usually around 2700MB in windows task manager when windows has loaded.
many thanks
.
Hi guys
What settings should I choose for prime blend test? Shall I just choose the default settings i.e 8k to 4096k and 1600MB ram or should I alter these for greater stress testing? I have 4GB ram (2x2GB) in dual channel mode and 'free' physical ram is always usually around 2700MB in windows task manager when windows has loaded.
many thanks
Ofcourse I can increase 'Free' physical ram to 3200MB. So guys any one? What settings should I use for Prime Blend test?