After realizing I did not place the intel stock hsf the right way because prime was giving me temperatures of 75 degrees after 2 minutes of blend stressing, I removed and placed it again and could move on.
Advice on my following questions is very much appreciated:
1. The vcore is set to 1.2 in the p5q bios, but cpu-z shows me 1.184 / 1.192?
2. Prime95 has been running for 3 hours now, and the cores are 60 / 62 degrees, this good and safe?
3. I chose the first option, Small FFTs, is this a good test and should I run it 24h?
4. If I wanted the temps to decrease would a Xigmatek HDT-S1283 and Arctic MX-2 thermal paste suffice?
I use DDR2-800, FSB is set to 400 and temps are monitored with SpeedFan, Real Temp and Core temp.
Oh and btw anyone know of a good 80x80x15mm cooler for my cm690? It's for the back of the motherboards, supposed to take off a few degrees when used as exhaust.
After realizing I did not place the intel stock hsf the right way because prime was giving me temperatures of 75 degrees after 2 minutes of blend stressing, I removed and placed it again and could move on.
Advice on my following questions is very much appreciated:
1. The vcore is set to 1.2 in the p5q bios, but cpu-z shows me 1.184 / 1.192?
2. Prime95 has been running for 3 hours now, and the cores are 60 / 62 degrees, this good and safe?
3. I chose the first option, Small FFTs, is this a good test and should I run it 24h?
4. If I wanted the temps to decrease would a Xigmatek HDT-S1283 and Arctic MX-2 thermal paste suffice?
I use DDR2-800, FSB is set to 400 and temps are monitored with SpeedFan, Real Temp and Core temp.
Oh and btw anyone know of a good 80x80x15mm cooler for my cm690? It's for the back of the motherboards, supposed to take off a few degrees when used as exhaust.
Thanks for reading,
Greets Rowan.
1. I have always noticed a little difference between what is being measured and what is being used. The best way to know for sure is to get a voltmeter and check it.
2. Yes 60/62 is right at the safe limit I think. But it can go higher before hitting the danger temps.
3. Couldn't tell ya.
4. I would want to say yes but I haven't got mine yet so I am not sure. I am almost 99.99% positive that it would lower it a good 10*C full usage.
And I have never found a fan to put on the back side of my Cm-690, but if you have enough fans pushing in air will naturally flow out of there. I have 60 cfm fans in my case in every fan place except on the side panel because of noise. I used the stock fans on there.
------------------------------CM-690 | E8400 | P5E X38 - Rampage Formula Bios | XFX 8800GT | 4 GB - HyperX 800Mhz 4-4-4-12 | XtremeGamer | Corsair 550W | Lite-On iHas120 | XP Pro SP3 - Blacked Out
Homebrew in a Dell Case | P4 HT 3.2Ghz | 865G | 1.5GB DDR | eVGA 7800GS | Audigy ZS
Reply to 98silvz71
1. That's normal fluctuation.
2. Fine, especially under Prime testing.
3. Any of the tests are fine. When you're done all your OCing, run a few hours, overnight should be fine.
4. Certainly.
------------------------------Phenom II X2 unlocked to four cores @ 3.8gHz,1.45V. 4GB DDR3-1600 Gigabyte 785g ATi Radeon HD4870 1GB
Run Folding@Home! Support Toms Hardware Guide, Team 40051!
CPU Buyer's Guide
Reply to smithereen
To tell you the truth I really can't tell a difference when I o/c'ed to 3.6 on mine. Speed wise with booting and loading stuff it all seems the same. But with things that are cpu intensive like say Bloons Tower Defense 3 I used alot of it in the higher levels and it could get pretty bogged down. Now with it o/c'ed I use less of the cpu and it doesn't get as bogged down. Same goes for things like WinRAR, it uses all of my 2 cores but once I o/c'ed it takes less time using my 2 cores. I'm sure that if I took it higher I might be able to see more of a difference. I kind of think it might be my ram. Because I tested out a buddies set of 1066Mhz stuff when he was having problems and I was o/c'ed to 3.6 it seemed alot snappier. Even when the ratio wasn't 1:1 with the FSb:ram freq. And btw, I have all that stuff enabled that you have off, except for the max cpuid. Thats how mine stays cool with it o/c'ed. Mine is still stable running on stock VID and it doesn't make a difference with it on or off on my computer.
------------------------------CM-690 | E8400 | P5E X38 - Rampage Formula Bios | XFX 8800GT | 4 GB - HyperX 800Mhz 4-4-4-12 | XtremeGamer | Corsair 550W | Lite-On iHas120 | XP Pro SP3 - Blacked Out
Homebrew in a Dell Case | P4 HT 3.2Ghz | 865G | 1.5GB DDR | eVGA 7800GS | Audigy ZS
Reply to 98silvz71
VID is the value that your CPU will set vcore to. If you can set it lower and still run, great.
"I've ran Prime for 11 hours now without errors/warnings. Should I lower the vcore again, if so, by how much?"
Once you are at the 3.6 GHz. You can try lowering your voltages until the system is no longer stable. At that point increase it a couple of steps.
----------
Overclocking since 1978 - Z80 (TRS-80) from 1.77 MHz to 2.01 MHz
I got it running at 3.6GHz with a vcore of 1.136 now. Prime95 has been going for 5 hours now, no errors When I set the vcore to 1.11 before, windows BSOD'd after a few seconds, so if this 1.13 is stable after 24hours, I'll just leave it at that.
Temps have also decreased to 56c degrees ^^ And jsc, thanks a lot for that link! Very helpful.
Just FYI, make sure you have round off checking enabled in prime95, or else some errors may not get returned and you system might still be unstable even though it passed prime95.
Also it might be a good idea to do a blend test as well, as that will test your memory. If you aren't actively changing the memory divider then whenever you increase the fsb your memory is going to get clocked higher as well, and could prove to be unstable. Prime95 small ffts is great for testing the processor, but it hardly tests the memory, so you could again end up with a successful prime95 test and yet still have an instable system (This actually happened to me just a few days ago)
I would still have to recommend running a blend test for at least a few hours. Chances are your memory is stable, but it is always better to be safe then sorry. I didn't bother running a blend once, and was getting a blue screen a day because of it.
After realizing I did not place the intel stock hsf the right way because prime was giving me temperatures of 75 degrees after 2 minutes of blend stressing, I removed and placed it again and could move on.
2. Prime95 has been running for 3 hours now, and the cores are 60 / 62 degrees
Last week I replaced the stock HSF with a Xigmatek Dark Knight, using Arctic MX-2 thermal paste. Load temps are now 42c I advise you getting one as well!
Did you try to lower the vCore, instead of leaving it at auto? Try lowering it.
I had an e6750 @ 3.7 with 1.4275 Vcore. My CPU temps were around 66C with OCCT/Prime 95 small FFTs. I still have a Xigmatek 1283 with a Scythe 133 CFM fan at half speed on my i7 920. My temps now are under 70C, but were close to 70C with my E6750. Only so much heat can be dissipated. If your temps are <60s with Prime95 small FFTs, then you must have your PC outside when it's -20C.