Abit IX38 Quad GT - Living Review II

[supershanks]

BSOD lol

seems stable at 350 @ 1:1.20

Try original fsb & up vdimm a bit , did you try prime?? when running primer blend make a note of the kind of fft it dails on.

[supershanks]

1st test :-
3600, 9x400@1:1.50 = 1200 DDR2 5,5,5,12



1st time on this board i've managed that memory speed.

[glowball]

The tip came from Johhny on Xtremesystems to be fair.
Ethel only plays online now as she's past pushing the wheel barrow full of cash back to the old folks home.
3600, 9x400@1:1.25 = 1200 DDR2 5,5,5,12 is really good news as that's what I run (unless x8 is better now) but as luck would have it - all my RAM is on it's way back to Crucial for replacement, just as a final BIOS is released. I will be watching this thread for comments.
What were the rest of your BIOS settings for the memory?
Any C1 errors?
Thanks and keep up the good work.

[karas]

Originally Posted by supershanks

Try original fsb & up vdimm a bit , did you try prime?? when running primer blend make a note of the kind of fft it dails on.

now it's 1hr20min im running the blend test with prime.. at 350fsb with vdimm 2.075v, seems stable as it is.. i had a peak on the PWM phase3 of 82°C but normally during the blend test is <75°C..

why trying the original fsb ?

at 365 it crashes just when blend starts.. at the very first test

[glowball]

Originally Posted by karas

...at 365 it crashes just when blend starts.. at the very first test

To find where the problem is...

Small fft's stresses the cpu
Large fft's stresses the memory

Blend doesn't tell you which is failing - unless off course you know exactly which blend test runs first.

[supershanks]

at 365 it crashes just when blend starts.. at the very first test

It's up to you up voltages & keep 365 or option 2 which you've goner with drop fsb.

Blend doesn't tell you which is failing - unless off course you know exactly which blend test runs first.

you do look again om your own picture

You can see that core 4 failed running large fft, This usually implies a memory error.For 365 - try upping vdimm back to 2.10v & rerun.
Results.txt is in the folder you run prime from.

i had a peak on the PWM phase3 of 82°C but normally during the blend test is <75°C..

Your better adding a fan if you can over the PWM cooler.

[karas]

Originally Posted by glowball

To find where the problem is...

Small fft's stresses the cpu
Large fft's stresses the memory

Blend doesn't tell you which is failing - unless off course you know exactly which blend test runs first.

i'd say.. small fft

EDIT: bsod with 355fsb too.. ok i'll try with vdimm @ 2.1v

[supershanks]

9x400 @1:1.25 DDR2 1200 - Official Bios 12

ABIT IX38 QuadGT

OC GURU:
 
CPU OPERATING SPEED................ - USER DEFINE
EXTERNAL CLOCK..................... -   400  MHz 
MULTIPLIER FACTOR.................. - x 9 
ESTIMATED NEW CPU CLOCK.......... -  3600  MHz 
DRAM SPEED (CPU/DRAM)........... - DDR2 - 1200 (1:1.25) 
PCI EXPRESS CLOCK.................. - 105MHz 
 
VOLTAGE CONTROL:
 
CPU CORE VOLTAGE................... - 1.4125V
CPU Core Voltage Offset............ - 0V
Target CPU Core Voltage............ - 1.4125V 
DDR2 VOLTAGE....................... - 2.075
CPU VTT VOLTAGE.................... - 1.30V
MCH 1.25V VOLTAGE.................. - 1.45V May be a tad ji??
ICH 1.05V VOLTAGE.................. - 1.12V
ICHIO 1.5V VOLTAGE................. - 1.50V 
DDR2 REFERENCE VOLTAGE............. Default
CPU GTLREF 0+2..................... - 68%
CPU GTLREF 1+4..................... - 67%
 
ADVANCED CHIPSET FEATURES:
 
DRAM TIMING SELECTABLE............. - MANUAL 
CAS LATENCY TIME (TCL)............... - 5
RAS TO CAS DELAY (TRCD)............. - 5
RAS PRECHARGE (TRP)......... .......... - 5
PRECHARGE DELAY (TRAS).............. - 12
REFRESH CYCLE TIME (TRFC)........... - 30
WRITE RECOVERY TIME (TWR)......... - 4
WRITE TO READ DELAY (TWTR)........ - 4
ACT TO ACT TIME (TRRD)............... - 4
READ TO PRECHARGE (TRTP)............ - 12
COMMAND RATE....................... - 2T 
PCIE Root Port Function............ - V1.0A 
PEG FORCE X1....................... - DISABLED 
INIT DISPLAY FIRST................. - PCIEX 
 
ADVANCED BIOS FEATURES / CPU FEATURES:
THERMAL CONTROL.................. - DISABLED
 C1E FUNCTION....................... - DISABLED 
 EXECUTE DISABLE BIT................ - DISABLED 
 VIRTUALISATION TECHNOLOGY.......... - DISABLED
 EIST FUNCTION...................... - DISABLED

Not checked stability yet

Any C1 errors?

Not since flashing to 12.

I've got an open mind whether it was the bios 12.09 or just the dmi pool screwing up again, as has been the case more than occasionally with this board.

[supershanks]

Back to Prime Large & small FFTs. THis shot shows prime stopped by user , but would be same if an error.

Here's one i made earlier



looking at the pic Tests are < 512 so are small FFTs. So the blend test was stopped on a small fft cycle which stresses cpu. You can further validate this by watching your cpu temps through the test small ffts cpu his at it's max, memory temps lower . large ffts the reverse.
Hense the variation between min & max :-


Luck

[vwgolfman]

Guys,
I need to ask a quick favour of the IX38-Quad GT possie.....

What is the difference between the on-board HD Audio and AC97 Audio?

The case into which I put this Mobo last night has two, 3.5mm jack sockets on the front. One for headphnoes and the other for a microphone.

The cable running from these sockets has 2 different terminations. Both are 8 or 10 way molex connectors and both would plug into the audio socket on the Mobo but I don't understand what it says about the audio in the Mobo manual.

The pinouts of the two molex housings are slighty different.

Which should I plug in and why?

Cheers guys