My cpuz is showing my computer running at 2800.3 mhz, multiplier at 8, my bus speed is at 350, rated fsb 1400.1 mhz and core voltage at 1.344v
My core temp 0.99.5 shows that core#0 54c with a 5% load
I got 2x 1gb ddr2 corsair and have it set in my bios at 4-4-4-12. Just ordered two more so it's gonna be a total of 4gb of rams.
With these settings, when I run the orthos, it only runs for about five mins then stops and gives me an error. I've been playing with the overclocking for at least 10+ hrs now. I'm not sure if I'm doing this correctly or somewhere in the bios I didn't do something right. I got a pretty cheap aftermarket cooler master hs, looks kinda like the stock one. Thinking about upgrading that when I get the chance. I would like to run it at least at 3Ghz.
Does running the orthos really matter? There are times where I've clocked at 3.04ghz at 1.425v and it was ok when I started up windows, but running that orthos, it never passes but the computer is still on.
Does anyone out there have a GA-965P DS3 board with the core 2 duo E6400 clocked at least 3GHz with 2gb of ddr2 corsair and stock heatsink that is stable. I would love to get the specs.
65 C is not a bad load temperature. Go into PC Health in your BIOS and set the thermal alarm to 70 C. However, 54 C is kind of high for an idling temperature.
Next, you have probably reached the limits of the stock heatsink.
There's a good chance that with 4 DIMM's, you will need to relax the memory timing and/or increase the RAM voltage a little to get the system to run stably.
Running a program like Orthos or Prime95 does two things. It puts a 100% load on the CPU - something that seldom happens for an extended period of time in normal operation - to check the core temps. It also checks how stable the system is by performing some really complex calculations.
People stress test for different periods of time: 2, 4, 6, 8, 12, or even 24 hours. I test my overclocks for 24 hours. I used to test for 12 hours, but one time I just let the system run overnight and next morning discovered that Prime95 failed at 14 hours. I regard system stability as very important - not necessarily critical because I am not running complex, life-or-death calculations on my systems.