Bad Axe 2 successful overclock with E6400
Posted: Sun Mar 25, 2007 1:12 pm
After quite some time since my last attempt, I decided to have another go at overclocking. My system is an Intel D975XBX2 ("Bad Axe 2") with an E6400 processor and 2 gig of Corsair DDR2-800 XMS2-6400 Xtreme Performance memory. CPU HSF is a Zalman CNPS9500AT with Artic Silver Ceramique. I first tried to overclock at the end of January this year (see http://forums.legitreviews.com/about7851.html), but had issues with maintaining the overclock across power cycles.
BIOS version 2674.
Room temp is 24C
Trial 1:
system BIOS settings:
CPU voltage override: 1.3375 (up one click from default)
Enhanced power slope: enabled
Memory frequency: 667
Memory timings: 5-5-5-12
Memory voltage: 1.92
MCH/ICH voltage override: 1.525 (up 1 click)
FSB voltage override: 1.25 (up 2 clicks)
with the FSB freq set to 300 (8x300=2400), Windows XP shows temps of 45C and 40C from IDU and TAT respectively. After running Orthos for 5 minutes, the temps climb to 58 (IDU) and 50 (TAT).
Under Linux (64-bit Gentoo), the system idles at 46C. Running two copies of mprime for 8 minutes results in the CPU temp as reported by lm_sensors/gkrellm maxing out at just about 60C.
Trial 2:
Changed FSB to 333: 8 x 333 = 2666
Idle temp is now 46 (IDU) and 41 (TAT); Linux reports 46.
Running Orthos under Windows XP: IDU reports 57-58, TAT reports 48-50 after 5 minutes.
Running 2x mprime under Linux for 52 minutes: 64C
Observations:
- As soon as I started overclocking, the system "lost" enhanced speedstep. I believe this was when I increased the CPU voltage. Many people turn this off manually as it causes problems apparently. This BIOS turns it off automatically?
- The TAT CPU load test gets the CPU the hottest of anything I've tried. I don't have intestinal fortitude enough to see just how high it will climb and click STOP.
- So far, I've only waited 45 minutes with power off to see if it survives cold starts. So far so good. Overnight tonight will be a better test.
Question:
Are these temperatures reasonable?
BIOS version 2674.
Room temp is 24C
Trial 1:
system BIOS settings:
CPU voltage override: 1.3375 (up one click from default)
Enhanced power slope: enabled
Memory frequency: 667
Memory timings: 5-5-5-12
Memory voltage: 1.92
MCH/ICH voltage override: 1.525 (up 1 click)
FSB voltage override: 1.25 (up 2 clicks)
with the FSB freq set to 300 (8x300=2400), Windows XP shows temps of 45C and 40C from IDU and TAT respectively. After running Orthos for 5 minutes, the temps climb to 58 (IDU) and 50 (TAT).
Under Linux (64-bit Gentoo), the system idles at 46C. Running two copies of mprime for 8 minutes results in the CPU temp as reported by lm_sensors/gkrellm maxing out at just about 60C.
Trial 2:
Changed FSB to 333: 8 x 333 = 2666
Idle temp is now 46 (IDU) and 41 (TAT); Linux reports 46.
Running Orthos under Windows XP: IDU reports 57-58, TAT reports 48-50 after 5 minutes.
Running 2x mprime under Linux for 52 minutes: 64C
Observations:
- As soon as I started overclocking, the system "lost" enhanced speedstep. I believe this was when I increased the CPU voltage. Many people turn this off manually as it causes problems apparently. This BIOS turns it off automatically?
- The TAT CPU load test gets the CPU the hottest of anything I've tried. I don't have intestinal fortitude enough to see just how high it will climb and click STOP.
- So far, I've only waited 45 minutes with power off to see if it survives cold starts. So far so good. Overnight tonight will be a better test.
Question:
Are these temperatures reasonable?