Reading Done - Time for Doing

[Sinque]

Over the past week or so i've been reading about overclocking and what to do and how to do things. So i've finally plucked up the courage to go a bit higher than an extra 20mhz on 200 My current rig is as follows:

AMD 64 3200+ Newcastle
GeIL 2x1GB PC3200 Value Dual Channel Kit CAS3
Asus K8N4-E Mobo ( Socket 754 )
GeForce 6800GS
3dAurora Chassis with two 80mm in-take and outtake fans.
Scythe Ninja Plus Rev.B Heatpipe CPU Cooler

So i've been messing about for a while now with bios etc and managed to get from 2.2ghz to 2.42ghz(220x11) nice and stable. But now i've managed to get to 2.53ghz(230x11). I know i've done some settings wrong somewhere and that some things are wrong but I just can't see where. Here is what it currently looks like:



So how close am I to blowing my computer?

To me the HT Link is too low, maybe i've read something wrong and maybe it's fine, and I think my RAM settings are wrong. I just changed them until XP booted

Going to see how long SP2004 lasts for.

 

[Sinque]

whats your temps like? for me command rate should always be 1T for performance.

Just ran SP2004 on 'Small FFTs' for 4minuts and 39seconds and my CPU got to the highest of 54c, barely though, and got this log:

Type: Small FFTs - stress CPU Min: 8 Max: 32 InPlace: Yes Mem: 8 Time: 15
CPU: 2530MHz FSB: 230MHz [230MHz x 11.0 est.]
02/01/2007 20:48 Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 800000 Lucas-Lehmer iterations of M172031 using 8K FFT length.
Test 2, 800000 Lucas-Lehmer iterations of M163839 using 8K FFT length.
FATAL ERROR: Rounding was 0.4995117188, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 4 minutes 39 seconds - 1 errors, 0 warnings.
Execution halted.

Will edit the post after Large FFTs and then Blend.

Edit 1: Large in-place FFTs ran for 1minute and 22second and got around 50C

Type: Large, in-place FFTs - stress some RAM Min: 128 Max: 1024 InPlace: Yes Mem: 8 Time: 15
CPU: 2530MHz FSB: 230MHz [230MHz x 11.0 est.]
02/01/2007 20:55 Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 1 minutes 23 seconds - 1 errors, 0 warnings.
Execution halted.

Edit 2: Blend ran for 2min 4seconds and got a max of 54C

Type: Large, in-place FFTs - stress some RAM Min: 128 Max: 1024 InPlace: Yes Mem: 8 Time: 15
CPU: 2530MHz FSB: 230MHz [230MHz x 11.0 est.]
02/01/2007 20:55 Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 1 minutes 23 seconds - 1 errors, 0 warnings.
Execution halted.

 

[Sinque]

Not that it makes a huge amount of difference (if any) but your HTT/LDT multiplier should stay at 4 until you get to 250 FSB, then reduce it to 3 to stop the overall speed HTT speed being greater than 1000MHz. Also, can you post a screenie of the 'SPD' tab from CPUZ?

My computer wouldn't boot if my my HTT was on 4 anywhere after 220 unless it was on 3. Will edit with SPD tab, one sec'

Edit:

 

[Sinque]

^^^

Drop your HT multi to x3 as well.

Jokester

Yeah I have, took me about 2hours to work that one out. Almost turned emo as I couldn't get it to work

set your ram timings to 3-4-4-8 just incasae thats whats limiting you

Done that and seeing what happens.

Edit : Command rate is now on 1t and all is fine.

 

[Sinque]

Hmm, I raised the VCore to 1.575 and my comp failed to load and so I reset the bios using the jumper. I instantly went into bios and overclocked once again back to everything above but like 1.55 again, i'm running SP2004 and i'm maxing out at 45C?! The only difference is my case side is off

SP2004 has been running Small FFTs for 7minutes now perfectly and still only 45C, i'm very confused? Surely the case side cannot make this much difference.

 

[Sinque]

That's normally a sign that air flow through your case is pretty bad.

Jokester

I'll take some photos, give me a minute.

Edit: Ok, sorry for the quality but it's from my phone.


There is the intake fan which blows air across my HDs and into my fan.(The blue thing on the front )



And here is my cooler, which also has a fan attached to the bottom which blows air up through the CPU and on the left are the two fans which pull air out of the computer.

 

[Sinque]

Nah, no need for photos, what case fans have you got?

Jokester

Too late

Edit: SP2004 been running for 16 minutes 48 seconds on Small FFTs @ 2530MHz (2530-2531) and still only 45C

 

[Sinque]

OT but I thought all A64 3200's had a 10x multi with a stock speed of 2Ghz?

Hmm, mine at stock was 11multi with 2.2GHz

 

[Sinque]

Probs just a difference performance index from socket 939 due to the lack of dual channel RAM capability on 754.

edit - beat me

Still stable?

Duration : 26 minutes 24 seconds

And at 45C still.

Edit :

Type: Small FFTs - stress CPU Min: 8 Max: 32 InPlace: Yes Mem: 8 Time: 15
CPU: 2530MHz FSB: 230MHz [230MHz x 11.0 est.]
02/01/2007 21:35 Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 800000 Lucas-Lehmer iterations of M172031 using 8K FFT length.
Test 2, 800000 Lucas-Lehmer iterations of M163839 using 8K FFT length.
Test 3, 800000 Lucas-Lehmer iterations of M159745 using 8K FFT length.
Test 4, 800000 Lucas-Lehmer iterations of M157695 using 8K FFT length.
Test 5, 800000 Lucas-Lehmer iterations of M155649 using 8K FFT length.
Self-test 8K passed!
Test 1, 560000 Lucas-Lehmer iterations of M212991 using 10K FFT length.
Test 2, 560000 Lucas-Lehmer iterations of M210415 using 10K FFT length.
Test 3, 560000 Lucas-Lehmer iterations of M208897 using 10K FFT length.
Test 4, 560000 Lucas-Lehmer iterations of M204799 using 10K FFT length.
Test 5, 560000 Lucas-Lehmer iterations of M200705 using 10K FFT length.
Torture Test ran 30 minutes 54 seconds - 0 errors, 0 warnings.
Execution halted.

It didn't get anywhere above 47C, maxed between 44 and 47.

Edit : Running large in place FFTs now.

 

[Sinque]

Just had a big ass problem with my keyboard and only just managed to fix it, resetting my bios was my attempt at fixing it. So when I did fix it I decided to up my fsb to 235 from 230. It works fine, idling temp at 35-40C, but sadly SP2004 doesn't agree. Errors practially instantly.

Anything i may need to change to get higher than 230?

Edit: I think my CPU needs more power but if I put it any higher than 1.55 it just dies

 

[Sinque]

I'd let it run overnight, say 8 hours worth to test its stability. You could always try a different higher voltage, sometimes boards dont like certain voltages or fsbs for whatever reason. Also, a bios update might be worth a shot. It could be that you've hit the limit of your chip though, and TBH 2.5ghz is alright for a newcastle.

The current is as solid as a rock and can't see it failing.

I'll be sure to try higher voltage, never thought of that.

I got the bios update and the flasher but i'm not sure how to do it, so I gave in

I just want to keep pushing as far as I can. It's addicting.