**Official ABIT IX38 QuadGT Review/Overclock/Guide Thread**

[Ace-a-Rue]

Thanks jcool..
I have another question that is does this mobo allow 4gb of ram to be run at 1066 or does it have to be run at 800 like most other boards? I have 4gb of dominators and was told to run them @ 800 Mhz but when i benchmark it gives worse results than when it had 2 gb @ 1066.

you have FSB to memory ratios...1:1 and 1:1.2 are 1333 mhz memory straps...the next two, 1:1.25 and 1:1.50, are 1066 straps...super pi is memory speed dependent besides raw power of the processor.

[jcool]

I think he wanted to know whether it's actually possible, not only in theory
I have been running my 4x1GB D9GMH at 1080Mhz, that is FSB 450 and using the 5:6 divider. You can also try reach that memspeed by using the 2:3 divider, since the 4:5 won't work that high (they still have to fix that).

[wilson]

Ok so if im at fsb 1422 what divider should i use? and should i run them at 5.5.5.18. instead of 5.5.5.15 and will i need more NB volt and im using 2.24v on my memory now is that enough???

[Supershanks]

for 24/7 running ram <=1000Mhz youi can use tighter timings 4,4,4,12.
>1000 you slacken them to 5,5,5,10~12 depending on your ram spec.

fsb 1422

is quad pumped /4 = 356 which is how people genarally talk about fsb
Available straps are :-
1:1.00 = 356 x DDR2 = 712Mhz
1:1.20 = 426 x2 = 852Mhz
1:1.25 = 444 x2 = 888Mhz
1:1.50 = 533 x2 = 1067Mhz your rated memory speed.

However not all straps are equally stable. That's really where you need to experiment yourself. You may need to increase Mch 1.25v voltage to help when running a higher strap.

The other laternative is to run a higher FSB, perhaps with a lower CPU muyltiplier, this will allow you to use a lower memory strap.

luck

[jcool]

Nice explanation supershanks, I think he got it now
@wilson: For your FSB, use the 1:1,5 (=2:3) divider to get your desired memspeed. Timings of 5-5-5-12/15 should be fine, just as Supershanks said.
And I don't think you need that much Vdimm, I was running 1080 with 2,1V so try to lower it a bit.

[wilson]

Ok cheers guys im now running @ 5.5.5.15. 1067Mhz with 2.3 divider and have passed two memtests. I am running my nb with 1.38v and sb with 1.15v and it seems to be stable . Is running them at this speed putting strain on the mem controller or does it not really matter if ive upped the volt abit??

[jcool]

I guess it's ok, since you're using aircooling the NB won't get too hot.

[Supershanks]

Where's your mobo at now jcool ??

[jcool]

You mean the two dead IX38? Well, they're both in RMA (didn't send the first ahead out of laziness... saved me the freight)
But somehow I'm not looking forward to a 3rd try... honestly what's the fun in having a single stage unit and not being able to use it properly for fear of a dying mobo

[Ace-a-Rue]

jcool, do you think your PWM cooling solution might have shorted out the digital PWM circuit?

[jcool]

Nope I don't. Since some other guys have the same problems now... And I triple-checked the PWM freezer, it does not touch anything except the PWMs.
And I don't think it's the PSU, not anymore. Not since my friend with the oscilloscope tested it and told me that it's perfectly fine, no ripple or spikes or whatever.
I think it's just >100A for a continuous period of time = goodbye PWMs, simple as that.

However, the guy that's developing these Mips freezers just sent a new prototype my way. This prototype is designed to use with thermal grease, not a pad. So maybe if I can keep the PWMs cooler with this prototype's help, the mobo won't die - I'll post some pics and let you know how it goes.
I am so not gonna step down from my OC because of a stupid mobo...

[Ace-a-Rue]

any chance of it wiggling or moving ever so slightly to touch the other circuits on either side of the digital chips?

[superbowtie]

anyone tried a e8400 yet?

[jcool]

any chance of it wiggling or moving ever so slightly to touch the other circuits on either side of the digital chips?

You could tie a rope onto the PWM freezer and play cowboy with the mobo and it still wouldn't move an inch so noooo way

[AMDfreak]

You could tie a rope onto the PWM freezer and play cowboy with the mobo and it still wouldn't move an inch so noooo way

[vanovich]

this is what i got out of this board so far ,http://service.futuremark.com/compare?3dm06=4931312
http://valid.x86-secret.com/show_oc.php?id=303631
just wierd comming back from a bad oc in biose

[ante_ante]

anyone tried a e8400 yet?

Hopefully i will get a E8400 tomorrow to test on the board...

[jcool]

I've got the E8400 but no more boards to test on...

[Supershanks]

anyone tried a e8400 yet?

yep
Intel E8400 Wolfdale Living Review verdict sofar, cpu looks promising, board is @*!! and the claim it's 45nm ready

Edit:
got to here 495x9 Despite this board. using shipping bios 11. really perverse after 2 months of bios development, you get a new cpu & the shipping bios works. Feel like i've entered the twighlight zone.

[superbowtie]

SuperShanks, have you tried the beta 6 or 8 version? I wouldn't even waste my time with 9 or final. can't be much longer till another beta comes out with the penyrn releases. Doesn't make any since how a newer version of a bios would be worst then previous. someone forget to save their work at Abit?

Talk about a bad luck. Mwave is like 205 for the chip but shipping takes it up to 218.61, Newegg is 219 with free shipping .

It seems The Egg has turned out to be someone's Nest egg. Can we say someone got Greedy.

[klataa]

Hello All

I was reading your posts and since the moderator allowed me to write i will give you some of my reflection about this board

I have Q6600@3600 stable on 1.41 offset 0.10 and its 24/7

http://valid.x86-secret.com/show_oc.php?id=303487

Ofcourse its nothing to compare with 4700mhz))))

I thing that this board is great to oc and stabbility with ours Q proccesor

My Q6600 can more than 3600 but im using the Infinity AC so i dont want to make him to hot in couple of days i will buy a WC and then ill try to oc more

I have a little question about this BEEP on startup.

When my pc is off and after couple hours i will turn it on, some times(not always) i can hear only ONE beeep. Is it normal??

Thanks to giving me advice on this forum - thats why i have bought quadGT x38

[ante_ante]

Now i have recieved my new E8400 ( Q744A761 ).

Running with final 12 bios and this is really strange. It keeps changing to 9.5 multi when i&#180;m in bios.

So i put a little lower clock and volts in bios and then i use uguru to raise FSB.

Testing 445x9 right now with 1.3125 and it&#180;s seems to work right now. Will try more What do u think about the bios, should i try another one? If so do they support 45nm cpus ??



Noticed that Cpuz reporting 1.448v but i have 1.3125 in bios

[Supershanks]

Changed Bios command line gets 45ns support in bios 12 using your command line cured screwed dmi & stopped the C1 erros.

Am running as smoothly as my ocing will allow now
Be interested to see if using your command line switches also get working multis

@superbowtie

It seems The Egg has turned out to be someone's Nest egg. Can we say someone got Greedy.

Yep theirs a few price gougers on this side of the pond, I managed to gert mine at a decent &#163;.

As i said above i think i'm looking good on 12 now after flashing using ace's command line. thanks. Be interested if it makes any difference to others who have had multi issues.

luck

[klataa]

I have a problem with stabbility at memory on 1:1 25 Any one know how to fix this issue?

Now i have on 1:1 20 and memory at 960 it would be nice to have at 1000

[Beastage]

I'm settling for 3200mhz on the cpu (with almost stock volts) but when I overclock my ram I get terrible memory performance, its Geil DDR2 800mhz 5-5-5-15 , I overclock it to 1000mhz (1:1.25 divider) at same timings but memory performance drop to 4000MB/s! , memtest86 does fine... what gives?