Sign in with
Sign up | Sign in
Your question

Q9450 OC, Prime95 application crash

Last response: in Overclocking
Share
August 12, 2008 4:13:21 PM

Here i go my own first topic...
After reading alot of posts on overclocking and reading the overclocking guide on this forum i began to overclock my system.(specs in signature). I was planning on a OC that would reach 3.2Ghz. I followed the instructions in de overclocking guide and eventually reached 3.2 so that my memory would be in sync. I started off with a Vcore of 1.35 on my OC and slightly tried bringing that down. When i reached 1.31875 my system seemed to be stable (1 hour of prime 95 large fft's testing without errors) at lower voltages i would get errors on 1 core withing 20 minutes or so. I decided to run a few hours on this OC setup and run prime95 overnight to confirm stability. The next morning i checked my system and saw that the prime95 application was crashed. There were no fatal errors on any core though. I thought i would raise the voltage a little to 1.32125 and then try again. The same thing happened again, after a few hours prime95 application crashed without any fatal errors. Does prime95 crashes because of an unstable overclock or could there be an other explanation??? I googled my problem but didn't came up with any god results that answer my question. All the prime95 testing was done with large fft's and the only voltages in the BIOS i touched was my Vcore.
August 12, 2008 4:31:53 PM

I would set memory timings on auto and give it the recommended voltage (2.1 volts?). That way you would know if it's a problem with cpu or memory.
August 12, 2008 4:39:25 PM

georgy said:
I would set memory timings on auto and give it the recommended voltage (2.1 volts?). That way you would know if it's a problem with cpu or memory.
.

Timings are set to the stock timings and the memory voltage is on auto. recommended voltages is 1.8.
I will try those settings and post back with a result.
Related resources
August 12, 2008 4:48:29 PM

First of all what is yiur cpu-s vid (realtemp can show it). For a 45nm cpu it requires quite a high voltage... I have a q9300 to reach 20 % overclock i had to rise it to 1.264 v. At quad core overclocking you may need to raise the north bridge voltage (1.40 is usually enough for 400 fsb) and a cpu vtt (fsb ovevolt mine required 1.2v from default of 1.1v). Without these voltages i was able to achieve a 10 % overclock max.
August 12, 2008 4:51:37 PM

ferencster said:
First of all what is yiur cpu-s vid (realtemp can show it). For a 45nm cpu it requires quite a high voltage... I have a q9300 to reach 20 % overclock i had to rise it to 1.264 v. At quad core overclocking you may need to raise the north bridge voltage (1.40 is usually enough for 400 fsb) and a cpu vtt (fsb ovevolt mine required 1.2v from default of 1.1v). Without these voltages i was able to achieve a 10 % overclock max.


VID = 1.2500
August 12, 2008 7:02:43 PM

Ok, not the best vid but it is ok. That high end board offers a lot of voltage changing options. You may need to do a little research on them before setting values. As i sad before for quad ocing you need to adjust the mobo voltages as well not only the vcore. (but if you have readed the overclocking guide in the forum than you probably already know that.)
August 12, 2008 7:21:22 PM

Bottom line is you need more voltage. Also are you disableing speedstep ect.....
August 12, 2008 7:37:49 PM

Prime crashes with that dont send message window (probably memory error) or it is crashing and displaying the core that stopped working, and other cores running (then more voltage needed for cpu and/or nb))
August 12, 2008 8:46:34 PM

Speedstep is enabled. and yes prime crashes with that "don't send message" window. I will disable speedstep and look at the memory voltages and then post back. And ferencster yes i knew about those NB and SB voltages but hoped i didn't have to change those to get a stable overclock.
August 12, 2008 9:20:48 PM

After changing the mentioned settings i got the same problem. Looks like i need to play around with the NB voltages a little bit. Since i didn't fully research that i will switch back to my stock settings and continue this o/c some other time. Thanks for the great input!
August 12, 2008 10:37:20 PM

RUn the memtest for errors, and make sure you are running your ram at default speed an timming.
August 13, 2008 1:13:43 AM

After 4+ sets of DDR2 800 ram, I never found a pair/set that was satisfied with less than 1.9 volts for stock, and 2.05 or better for OC and Prime 95/Orthos stability.

Northbridge voltage on the intel boards is a lot less voltage hungry than the nvidia stuff(prior to 790i).
August 13, 2008 1:19:51 AM

Yes, I didn't know OCZ made any RAM with a spec voltage of 1.8V! Are you sure your OCZ is spec'd at 1.8V?

August 13, 2008 8:05:23 PM

X5 - I think you've got the wrong specs there. You list your RAM as "Gold XTC", while the RAM in the link is "Gold GX XTC". Also, the RAM in the link only comes in 512MB and 1GB modules, while you list yours as 2GB modules.
What is the actual model number of your RAM? I found some 2x2GB OCZ Gold RAM on newegg, model OCZ2G8004GK (http://www.newegg.com/Product/Product.aspx?Item=N82E168...), but its spec voltage is 2.0V.


August 13, 2008 8:27:53 PM

Mondoman said:
X5 - I think you've got the wrong specs there. You list your RAM as "Gold XTC", while the RAM in the link is "Gold GX XTC". Also, the RAM in the link only comes in 512MB and 1GB modules, while you list yours as 2GB modules.
What is the actual model number of your RAM? I found some 2x2GB OCZ Gold RAM on newegg, model OCZ2G8004GK (http://www.newegg.com/Product/Product.aspx?Item=N82E168...), but its spec voltage is 2.0V.


You're completely right. The online shop where i bought the memory listed that link so i assumed those were the right specs.
I will try prime again with the correct voltage and see what happens. Thanks for your input !

August 13, 2008 9:09:13 PM

Just had a run with 2.0 memory voltage but got the same problem. since i the memory allows it i am trying another run with 2.1. Let's pray thats the key.
August 13, 2008 9:15:12 PM

Still no luck :??: . do you guys recommend trying 2.2v on the memory or do i need to start looking for other options ?
I did memtest a few days ago for 2 hour without errors. Maybe i should do it again and let i run a bit longer ?
August 17, 2008 3:01:19 PM

Just a little bump up
a b à CPUs
August 17, 2008 6:45:57 PM

Try some experimenting with the CPU VTT voltage. It might be called FSB Termination Voltage in your BIOS.
!