q6600 9x400 @1.46v ok for 24/7 use?

brainhulk

Diamond Member
Sep 14, 2007
9,376
454
126
do u think thats too much for this proc and system. I was using 8x400 but wanted to see if this was ok. I'm getting 57c hottest core when econding divx files.

q6600 GO 400x9 tuniq tower with ac air funneled in when hot ambient
MSI 8800ultra
1gb x2 crucial ballistix
abit ip35-pro
75g raptors x2 raid 0
antec truepower quatro 850w PS
 

kenrippy

Golden Member
Sep 3, 2002
1,763
0
0
not even 8-10hrs, give it 15-20mins and you'll see the peak temps.

you should be good though. i've got mine at the same setting & 1.4775v in the bios. that's what it takes to get my G0 stable under prime tests. same board, cooler and ram as you too. (4x1gb ballistix)

just keep C1E & EIST enabled in the bios and you'll be a happy camper. no need to run at 3.6ghz & 1.46v even when it's idle.
 

jonmcc33

Banned
Feb 24, 2002
1,504
0
0
Originally posted by: adairusmc
what temps do you get after 8-10 hours of prime95 torture testing?

What makes you think there's a difference in temp from 10 minutes and 10 hours? Use Orthos and not Prime95.
 

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
21,067
3,574
126
Originally posted by: jonmcc33
Originally posted by: adairusmc
what temps do you get after 8-10 hours of prime95 torture testing?

What makes you think there's a difference in temp from 10 minutes and 10 hours? Use Orthos and not Prime95.

dude... u obviously dont know your stress software

Prime95 25.3 > Orthos



Prime95 allows you to run quadcores on them as long as its 25.3 or higher under 1 instance
Like so:
http://i125.photobucket.com/al.../aigomorla/3.73ghz.jpg

Orthos you need to run 2 instances and then specifiy in windows which core to be run on which application. Orthos was originally intended for use on dual cores and not quadcores.

Dont confuse the OP please.



And you really need to load her up on Prime. The reason is because under worst case situations, prime will push your cpu at that mark. Then you can evaluate what the max heat output of your quad will be and work accordingly to it.


 

adairusmc

Diamond Member
Jul 24, 2006
7,095
78
91
Originally posted by: jonmcc33
Originally posted by: adairusmc
what temps do you get after 8-10 hours of prime95 torture testing?

What makes you think there's a difference in temp from 10 minutes and 10 hours? Use Orthos and not Prime95.

I only mentioned that because he made no mention of testing the stability of his OC (so I assumed he did not).

I know there is no difference.

Prime95 25.3 or higher > Orthos
 

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
21,067
3,574
126
Originally posted by: adairusmc
Originally posted by: jonmcc33
Originally posted by: adairusmc
what temps do you get after 8-10 hours of prime95 torture testing?

What makes you think there's a difference in temp from 10 minutes and 10 hours? Use Orthos and not Prime95.

I only mentioned that because he made no mention of testing the stability of his OC (so I assumed he did not).

I know there is no difference.

Prime95 25.3 or higher > Orthos

longer term = more cycles of small and large fft. Small FFT = cpu torcher test. Large is usually the mem side.

Blend which is a mix of both is good for long term stability because it cycles though both. I personally do at least a 6+Hour prime to be sure my rigs are absolutely WCG stable.

Also on quads theres a nasty vdroop. Remember your feeding 2 dies and not 1. So sometimes it will pass for 2 hours and then fail at hour 3-4 which would mean you would need to up the voltage a bit.

Jonmcc33 if you dont know your stuff, you really shouldnt be advising.

There is a temp difference between 10min and 10 hours. Because on water cooling units, dependant on volumn of water, it could take up to 2 hours for the water to reach its max Delta T.

Same can be said on really large class sinks taking longer then 10 min to reach max DeltaT.
 

TC91

Golden Member
Jul 9, 2007
1,164
0
0
i prefer doing 24hrs of small fft's for stability testing. for voltage, i believe it is somewhat safe (don't quote me on this) to have the voltages +10% over stock voltage. so say if the stock volts is 1.35v, then 1.35v x 1.1 = 1.485v max. try to have the temps 20c under the tjunction at the very minimum.
 

firewolfsm

Golden Member
Oct 16, 2005
1,848
29
91
Temps under 70C load are safe, you can usually go a little higher but I don't like it
 

GFORCE100

Golden Member
Oct 9, 1999
1,102
0
76
Originally posted by: kenrippy
not even 8-10hrs, give it 15-20mins and you'll see the peak temps.

you should be good though. i've got mine at the same setting & 1.4775v in the bios. that's what it takes to get my G0 stable under prime tests. same board, cooler and ram as you too. (4x1gb ballistix)

just keep C1E & EIST enabled in the bios and you'll be a happy camper. no need to run at 3.6ghz & 1.46v even when it's idle.

Your Q6600 must have a low VID or 1.2500 or less if it gets by with 1.475V in the BIOS. These boards have a large vdroop so uGuru must be showing 1.38V under load or so at 3.6GHz.

Am I right? :)
 

kenrippy

Golden Member
Sep 3, 2002
1,763
0
0
Originally posted by: GFORCE100
Originally posted by: kenrippy
not even 8-10hrs, give it 15-20mins and you'll see the peak temps.

you should be good though. i've got mine at the same setting & 1.4775v in the bios. that's what it takes to get my G0 stable under prime tests. same board, cooler and ram as you too. (4x1gb ballistix)

just keep C1E & EIST enabled in the bios and you'll be a happy camper. no need to run at 3.6ghz & 1.46v even when it's idle.

Your Q6600 must have a low VID or 1.2500 or less if it gets by with 1.475V in the BIOS. These boards have a large vdroop so uGuru must be showing 1.38V under load or so at 3.6GHz.

Am I right? :)

yep, you're right. my VID is 1.2375v and under load CPU-z & uguru shows between 1.38-1.41v
 

Acanthus

Lifer
Aug 28, 2001
19,915
2
76
ostif.org
Originally posted by: kenrippy
Originally posted by: GFORCE100
Originally posted by: kenrippy
not even 8-10hrs, give it 15-20mins and you'll see the peak temps.

you should be good though. i've got mine at the same setting & 1.4775v in the bios. that's what it takes to get my G0 stable under prime tests. same board, cooler and ram as you too. (4x1gb ballistix)

just keep C1E & EIST enabled in the bios and you'll be a happy camper. no need to run at 3.6ghz & 1.46v even when it's idle.

Your Q6600 must have a low VID or 1.2500 or less if it gets by with 1.475V in the BIOS. These boards have a large vdroop so uGuru must be showing 1.38V under load or so at 3.6GHz.

Am I right? :)

yep, you're right. my VID is 1.2375v and under load CPU-z & uguru shows between 1.38-1.41v

Can i trade you CPUs? :D
 

Assimilator1

Elite Member
Nov 4, 1999
24,155
520
126
aigomorla
Blend which is a mix of both is good for long term stability because it cycles though both. I personally do at least a 6+Hour prime to be sure my rigs are absolutely WCG stable.

6hrs is nowhere near enough ,you need to do 24hrs to be as near certain as possible of 100% stabilty.Several times I've had it fail at 18hrs.
Oh & btw 24hrs that's also recommended by Primes author(writer?).
 

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
21,067
3,574
126
Originally posted by: Assimilator1
aigomorla
Blend which is a mix of both is good for long term stability because it cycles though both. I personally do at least a 6+Hour prime to be sure my rigs are absolutely WCG stable.

6hrs is nowhere near enough ,you need to do 24hrs to be as near certain as possible of 100% stabilty.Several times I've had it fail at 18hrs.
Oh & btw 24hrs that's also recommended by Primes author(writer?).

your absolutely correct.

I am just speaking out of my personal favorate. The longer the better, however, ive had fairly good luck at 6h+ on primes. Usually 7h is where most people will take it as offical.
 

adairusmc

Diamond Member
Jul 24, 2006
7,095
78
91
Originally posted by: aigomorla
Originally posted by: Assimilator1
aigomorla
Blend which is a mix of both is good for long term stability because it cycles though both. I personally do at least a 6+Hour prime to be sure my rigs are absolutely WCG stable.

6hrs is nowhere near enough ,you need to do 24hrs to be as near certain as possible of 100% stabilty.Several times I've had it fail at 18hrs.
Oh & btw 24hrs that's also recommended by Primes author(writer?).

your absolutely correct.

I am just speaking out of my personal favorate. The longer the better, however, ive had fairly good luck at 6h+ on primes. Usually 7h is where most people will take it as offical.

I never test over 12 hours myself, and I never will.

Most of the time I cut it off after 8 hours if there are no errors.
 

KIAman

Diamond Member
Mar 7, 2001
3,342
23
81
Bah, not only do I torture test my cpu for 5 weeks, I also throw it in the water, kick it around and set it on fire.

Stability for the win!
 

kenrippy

Golden Member
Sep 3, 2002
1,763
0
0
Originally posted by: KIAman
Bah, not only do I torture test my cpu for 5 weeks, I also throw it in the water, kick it around and set it on fire.

Stability for the win!

ha! love it

hey, where's the OP anyway?
 

brainhulk

Diamond Member
Sep 14, 2007
9,376
454
126
/peeks from behind rock

I'm lurking because i want to hear what more peeps have to say about torture testing. I've never done it on my oc's. If it passes 3dmark, I can play wow/and crysis demo for hours, and encode divx files (using xvid4psp, x264) for hours im happy for now...

but anyways I've been using 9x400 @ 1.46v and I love it. the ac helps alot with temps.
but i dont know if i want to nuke the proc with prime 95 for 8 hours.
 

kenrippy

Golden Member
Sep 3, 2002
1,763
0
0
Originally posted by: brainhulk
/peeks from behind rock

I'm lurking because i want to hear what more peeps have to say about torture testing. I've never done it on my oc's. If it passes 3dmark, I can play wow/and crysis demo for hours, and encode divx files (using xvid4psp, x264) for hours im happy for now...

but anyways I've been using 9x400 @ 1.46v and I love it. the ac helps alot with temps.
but i dont know if i want to nuke the proc with prime 95 for 8 hours.

i say if it will go 2-3hrs then that's all i need. i won't be encoding a video for more than 2hrs straight anyway. i don't want to run my cpu @ 71c for 24hrs straight :roll:
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
Originally posted by: KIAman
Bah, not only do I torture test my cpu for 5 weeks, I also throw it in the water, kick it around and set it on fire.

Stability for the win!
don't forget to chinese water torture test it. that got my last 3 cpus...

 

VirtualLarry

No Lifer
Aug 25, 2001
56,585
10,225
126
Prime95 small FFTs for 24hr, and then blend for 24hr, only then would I consider it stable. (Don't forget a bootable Memtest86+ run for 24hr too.)
 

terentenet

Senior member
Nov 8, 2005
387
0
0
Do you guys really do all that testing? I mean.... 24h this, 24h that and another 24h to go along. That's 3 days of testing just to make sure your computer is stable.
What are you building there? NASA computers? Do you have critical tasks to do with your computer (your life depends on them)? I for one don't mind if my computer locks once a week. I have Nvidia to blame for sh***ty drivers and the electric company for interferences in the power lines.
When I build my computer, I do a couple of 3d marks 2k6 to make sure it all goes well as a whole, then 2-3hrs of prime95 on all 4 cores.
That's 4hours of testing, out of 72.
The other 68h I use for gaming and enjoying the computer I built.
 

F1shF4t

Golden Member
Oct 18, 2005
1,583
1
71
Originally posted by: terentenet
Do you guys really do all that testing? I mean.... 24h this, 24h that and another 24h to go along. That's 3 days of testing just to make sure your computer is stable.
What are you building there? NASA computers? Do you have critical tasks to do with your computer (your life depends on them)? I for one don't mind if my computer locks once a week. I have Nvidia to blame for sh***ty drivers and the electric company for interferences in the power lines.
When I build my computer, I do a couple of 3d marks 2k6 to make sure it all goes well as a whole, then 2-3hrs of prime95 on all 4 cores.
That's 4hours of testing, out of 72.
The other 68h I use for gaming and enjoying the computer I built.


I would say 8 hours is good enough indication to make sure the comp is stable, memtest for a couple of hours also does not hurt. If you want stability don't overclock, but even that does not fully guarantee a stable pc.
 

jeffw2767602

Banned
Aug 22, 2007
328
0
0
I just start up Prime Small FFT's and hit the sack. If I wake up and my computer has not caught fire and is still chugging away, then thats good enough for me.
 

lektrix

Golden Member
Aug 9, 2003
1,174
0
76
Originally posted by: terentenet
Do you guys really do all that testing? I mean.... 24h this, 24h that and another 24h to go along. That's 3 days of testing just to make sure your computer is stable.
What are you building there? NASA computers? Do you have critical tasks to do with your computer (your life depends on them)? I for one don't mind if my computer locks once a week. I have Nvidia to blame for sh***ty drivers and the electric company for interferences in the power lines.
When I build my computer, I do a couple of 3d marks 2k6 to make sure it all goes well as a whole, then 2-3hrs of prime95 on all 4 cores.
That's 4hours of testing, out of 72.
The other 68h I use for gaming and enjoying the computer I built.

3 days out of a possible 900 that you'll use gaming, BIG DEAL.

If you want stability, you obviously have to test it under harsh conditions.