Orthos Blend Test for at least 12hrs is good for me, but i doing longer on this CPU for the fun of it and right now i am at 15hrs 19min and still going and i may shut her down at 16hrs... maybe.
Printable View
Orthos Blend Test for at least 12hrs is good for me, but i doing longer on this CPU for the fun of it and right now i am at 15hrs 19min and still going and i may shut her down at 16hrs... maybe.
i got intel Q6600 G0 last year i hit 3.2 ghz on air run 24 hrs stable with prime95 after one year my cpu wont stable all setting same anyone know how to fix this should i reseat heatsink put new AS5 or get new water cool system i want my cpu hit 3.2 ghz again
I already posted in your thread no need to keep posting same question in other threads... :)
Maybe its degraded.Quote:
Originally Posted by lioncool
Quote:
Originally Posted by lioncool
might have to give some more vcoreQuote:
Originally Posted by fatguy1992
Guys he has his own dedicated thread http://www.xtremesystems.org/forums/...hp?t=200732for this question..... :)
I posted in there and here about his issue and asked all that you have already and even i am waiting for his response..... :) :)
*snip* Look below...
You might want to take a glance at the Q6600 thread in the Intel section, since this wouldn't really be considered extreme overclocking. The Overclocking Q6600 thread has helped many people including myself get my cpu past 3.4GHz
So I got one of these Q6600 G0 CPUs lately and I've been thinking of OCing it to 3.6GHz. This is proving to be quite a trip as right now I have vcore set to 1.48750 in bios and it's still unstable. I have put 400x9 and I also have the newest F11 bios for GA-P35-DS3P rev 1.0. Before OCing more I had a stable OC at 3.2 (356x9, vcore 1.3850 in bios). Any ideas or help? Should I just ramp up the vcore a little bit more?
Q6600 G0 1.3125VID
Tunic Tower lapped a bit
Gigabyte GA-P35-DS3P REV 1.0 BIOS F11
2X2GB OCZ PLATINUM XTC DDR2 800MHz @ 5-5-5-15 907MHz
Current testing: 378x9, 1.41250Vcore in bios
http://img231.imageshack.us/img231/2081/43833110js5.png
Almost 2 hours and still going strong... Any tips for getting this to 3.6GHz? Although the performance difference between 3.4 and 3.6GHz (or 3.2 and 3.6) should be quite minimal.
D-TeK FuZion H20
Q6600 37A 1.45 (bios) 1.4 under load
Abit Ip-35 Pro Bios ID:14
8GB G.Skill 4-5-5-15
It's Already 24/7 Stable Which Was A Few Months Ago, Decided To Run It Again For A Screenshot
http://img185.imageshack.us/img185/5...7a14hs7.th.png
Wow this is insane, thanks to Jor3lBR for the chip ;).
Vcore: 1.276 measured with DMM, +1h small ffts:
http://img384.imageshack.us/img384/6649/insaneip1.jpg
Insane it is! WR for lowest vcore to get 3600Mhz Prime95 stable!Quote:
Originally Posted by nfm
What's your mobo?
What more could you ask for?! :up:
Syvere, that's a tough board to overclock the quads on. Take the nb sink off and put some good TIM in there and then use a small fan on the sink after. If you have some small sinks for the mosfets use them. Sometimes I just use old sb sinks and cut them into tiny bits to use for mosfet sinks on cheap boards. How much are your fsb and mch overvoltage control set at?
@nfm: nice oc! :up: what's your bios settings for that? i might wanna try it out :D
DFI LP P35 since it is my favorite :).Quote:
Originally Posted by Jor3lBR
@emoners
VTT 1.346v
GTL Refs:
CPU 1/3 0.894v 66.42%
CPU 0/2 0.896v 66.57% (Core 0 is weaker)
NB 0.895v 66.49%
http://img380.imageshack.us/img380/2843/p1050633bk4.jpg
http://img122.imageshack.us/img122/3818/p1050621pp3.jpg
http://img83.imageshack.us/img83/6347/p1050629yd8.jpg
http://img83.imageshack.us/img83/7115/p1050630ej2.jpg
All measured via DMM, different PSU = different GTL values. I needed above vNB and VTT for DDR1200, this chip can run this on stock vNB and 1.3v VTT.
I'm on old OCZ PowerStream 520W.
Really nice OC, but not LV WR ;)
I've seen 2 or 3 similar CPUs in the German HWluxx forums, incidentially on the same mobo. The DFI LT/UT P35 rock.. they're even better than my X38 except the X38 clocks 4x1gb a bit better.
Now try how low you can go with 4Ghz, mine needs around 1,42V (GTL disabled tho :wasntme: )
Well real vcore with dmulti I was getting 1.26v under load in Prime95 and to be honest I never saw a single screenshot of under 1.29vcore Prime95 10hrs blend for 3600Mhz and Q6600 so I'm pretty sure 1.28v is WR or close to it :up:Quote:
Originally Posted by jcool
Without a Phase Change temps will be above 60oC and consequently you loose stability for testing thats why he will prob. need more vcore.
A good WC system could take nvm to 4Ghz around 1.43 or 1.44 with slick GTL settings.
0, I haven't fiddled with those at all yet.Quote:
Originally Posted by loonym
4Ghz will need at least 1.435v from what I'm measuring, and lower ambient temp. The 3.6Ghz was done with 25.5C ambient. Yes this mobo is awesome, I did 10 loops of blend using same settings with my avg. 2x1024MB D9 Ballistix at 636Mhz, 2.193v, vNB 1.628v, 5-5-5-15 and really good sub-timings. I had to play with skews/delays on the memory.Quote:
Originally Posted by jcool
What was the ambient temp @ 3.6Ghz? I'm afraid I can't go that low. Here's what I got with 21C ambient:Quote:
Originally Posted by Jor3lBR
Vcore: 1.359v real
http://img403.imageshack.us/img403/7609/325ghzej7.jpg
My cpu prime stable at 8x460MHz but would like to know if someone can post some info on bios setup for FSB @ 470MHz or over. Trying to find out what's max FSB at x8 multi for the cpu, need like 1.5V NB for current setup.
Tnx. :wave:
With all due respect to you all, prime is :banana::banana::banana::banana:, pure :banana::banana::banana::banana:, been prime stable @ 3.6 ghz for 48 h running and running a single try at intelburntest 1.6 about 5-10 or 15-20 min crashed the app (or shows calc errors if you chose to turn off error detecting), therefore i laugh on ur faces when i see prime shots which is a joke.
But i imagine ur comps would also be intelburntest stable if ur vdroop is excellent, mine is not, drops 0.05v even with Load Line Calibration (P5K PRO)
This kiddie attitude will get you pretty far... I've posted numerous times that there is NO bulletproof test version... you can have linpack stability yet crash on prime, OCCT or a 3dmark,... this is why I suggest you to run numerous apps and verify then... and the best stability is daily usage... if it doesn't cap out or produce errors or acts weird it's good enough...Quote:
Originally Posted by eagzor
Some prefer 10 runs of linpack others go as crazy as 100 runs, some do 2 hours of prime some like you prefer 48 hours... none of them methods will give you 100% accuracy as that is just not achieveable... but each prefers his own testing method and I respect that...
Get your act together and clear ya head... then come back or stay where ya were at...
If im in a hurry I run prime for 30mins, if not 2 hours max...Quote:
Originally Posted by Leeghoofd
If I can get prime to be stable for two hours, my computer is stable, because running it daily with what usage I give it, it will never crash, never has yet anyway.
It seems weird you would bother leaving your computer running for 48 hours under full load, what an electricity bill you must have.
And as leeghoofd said, what a stupid attitude, not to mention primes fairly often the standard for stability tests. But really the most stable computer is one that doesnt crash for daily use, and no single test can 100% guarentee that.
My apologies, it was not my intention to start a flame war, i never said there was ever a bulletproof test.
All i'm trying to say is that prime is fully replaced by intelburntest since it does a good job upping the heat by many degree from prime, and the vdropp is dropped even further than prime, so why would you even bother use prime? That at least is my opinion.
EDIT:
And maybe i should elaborate my previous post, indeed i stress tested prime 48h, but after the countless failed intebultests, i ran prime again, maybe a couple hours only. to my surprise it was fully stable, and then i reran intelburntest, it failed. And to ensure it, i did that one more time, same results, prime succeeded, intelburntest failed. To get it stable i raised my vcore, both was then stable.
I have used prime for quite some time, maybe just like you, it's standard to test your overclocked computer with prime just as you say. But now something better is out, and seeing people taking prime-shots and with the experience i had with intelburntest for overclocking (not a single computer build with it has been crashing even once) makes me mad. And before with prime-stable overclocks, sometimes but not always it just crashed and i had to start looking at voltages and and temps all over.