My 1st attempt at overclocking (with pic)

martianrobot

martianrobot

martianrobot

Associate
Joined
1 Jan 2006
Posts
252
Location
Sheffield
Okay, heres my first proper proven attempt at major overclocking of my new rig... well, my first attempt at major overclocking of any rig actually:

db2917debc.gif


I didn't go the full 12 hours with Prime95 'cos I want to go to bed and not make the ozone hole any bigger, but I think it's pretty stable. No, I was only running one instance but whilst doing that I was surfing the net and playing music and videos in Media Centre. Yes, I underclock my X1800XT when in 2D mode so I can use less power and slow down the fans. Ignore Speedfan's 7000rpm CPU fan figure though, my Thermaltake Big Typhoon 120mm fan is actually running at a very quiet 1300rpm.

Am getting about 20.4 in SuperPi in case you are wondering.

Okay, so it's not up there with The Asgard's amazing 3.6ghz overclocking efforts of the same chip, but I'm pretty happy as a noob and think it's worked out nice with such an easy 52% overclock. I wasn't getting anywhere until I upgraded to the new DS4 F3 BIOS though.

For anyone noobish out there with the same kit (see sig) who wants an easy, safe but otherwise impressive overclock, heres my settings:

Boot into BIOS, then press Ctrl + F1 to enable hidden settings then go into MIT.

Set CPU multiplier to 7 x 400

Set memory multiplier to 2.00

Disable CIA2

Set memory timings as follows:
CAS Latency - 4
RAS to CAS delay - 4
RAS Precharge - 4
Cycle Time (Tras) - 12

All other settings in MIT to Auto or Normal

You also might need to disable the following in the Advanced BIOS section:

CPU Halt CIE
CPU Thermal Monitor
CPU EIST
Virtualisation or whatever its called

And you might need to disable FanSpeed Control Method in PC Health section too.

Save and reboot and crossfingers ;)

Hope this works for you, and if any experts out there have noticed a booboo in what I've done, please post.

Ta.
 
Last edited:
mightynimrod said:
Well done matey, always good to get free speed :)

Now go for 3ghz
Click to expand...

Hmmm... cheers mate, but that involves messing about with other voltages... harum scarum!

I am quite happy with what I got - its what I was wanting to aim for when taking delivery of my new Geil RAM. An extra 200mhz doesn't really warrant the effort to me... I'd rather spend my time twatting goblins in Ayleid ruins at 60fps maxed out.

Now if someone could tell me how to get up to the 100% overclock without blowing my system up or forking out for water cooling that might be worth my while ;)

Sorry for large picture posting - meant to set it up as a linked image rather than inline.
 
martianrobot said:
Okay... am trying that... how long should I run that for? Leave it overnight?
Click to expand...


Yeah, if it fails then you can always mess with the settings tomorrow :)

Priming can be a right pain, but always better to have a stable system.

Most people prime for 12 hours or so, i only used to prime for roughly how long i'd sit there gaming for. 4-6 hours.
 
martianrobot said:
Okay... am trying that... how long should I run that for? Leave it overnight?
Click to expand...
Your going to do it infrequently, unless your an habitual overclocker, might as well do it right. I'd say at least as long as you use the rig, 8hrs, 12hrs is better. Maybe 2 times, once for small ffts (cpu) and once for blend (cpu+memory).
 
Last edited:
Oops! That didn't get very far:


Code: 
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 1656 Time: 15
CPU: 2800MHz FSB: 400MHz [400MHz x 7.0 est.]
CPU: 2800MHz FSB: 400MHz [400MHz x 7.0 est.]
17/08/2006 00:55 
Launching 2 threads...
1:Using CPU #0
2:Using CPU #1
1:Beginning a continuous self-test to check your computer.
2:Beginning a continuous self-test to check your computer.
1:Press Stop to end this test.
2:Press Stop to end this test.
1:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
1:Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
2:Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
1:Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
2:Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
1:Test 4, 4000 Lucas-Lehmer iterations of M19174369 using 1024K FFT length.
2:Test 4, 4000 Lucas-Lehmer iterations of M19174369 using 1024K FFT length.
1:Test 5, 4000 Lucas-Lehmer iterations of M18874369 using 1024K FFT length.
2:Test 5, 4000 Lucas-Lehmer iterations of M18874369 using 1024K FFT length.
1:FATAL ERROR: Rounding was 0.5, expected less than 0.4
1:Hardware failure detected, consult stress.txt file.
1:Torture Test ran 8 minutes 59 seconds - 1 errors, 0 warnings.
1:Execution halted.
2:Torture Test ran 8 minutes 59 seconds - 0 errors, 0 warnings.
2:Execution halted.

So much for running overnight... hardly had time to clean my teeth!

Too tired to sort it out now... any suggestions posted here gratefully received in the morning :)
 
Nazca said:
Increase your VDIMM to 2.2-2.3V.
Click to expand...

Okay, got sidetracked and still not in bed yet (damn these forums are addictive!). Haven't changed the VDIMM yet, but I set the SP2004 stress test running again and its done 45mins so far without problems. Not sure what made it drop out so quick before.

Will leave it running overnight to see what happens.

cheers
 
iBot said:
interested to know how it did :)
Click to expand...

Well, just taken a look and so far its done 8 hours and 59 minutes without falling over :)

Not sure why it crunked after 8 mins and 59 secs on my first go (bizarre 8:59 coincidence there) ... perhaps something else I was doing interfered with it?

Will leave it for a total of 12hrs and see what happens, then post a screenie if successful.
 
Ive had probs like that - dual prime runs fine but do something on comp at the same time and it would fail. If you use orthos then set priority to 9 (not 10 as system does its own stuff like virus scans etc). Also use small fft's test as that will stress cpu. At the moment your stressing ram as well so if it fails your not sure which ones at fault. Test your ram seperately
 
ted34 said:
Ive had probs like that - dual prime runs fine but do something on comp at the same time and it would fail. If you use orthos then set priority to 9 (not 10 as system does its own stuff like virus scans etc). Also use small fft's test as that will stress cpu. At the moment your stressing ram as well so if it fails your not sure which ones at fault. Test your ram seperately
Click to expand...

Good point.

Hmmm... my orthos priority is set at 1. Doh! Should've read up on this before setting it going.

Mind you, both CPUs are being used 100% and I'm still able to do stuff without lag. Nice chips these Core2s :)
 
You must log in or register to reply here.
Back
Top Bottom