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

[ScubaMob]

Hi all!
I have finally gotten serious about OCing my q6600 with this board and I have found a couple of things:

I can reach 3.33ghz completely stable with vcore @1.3875 and 0.0 CPU offset. I have to have a target voltage of at least 1.4125 just to sniff 3.4ghz I'm currently running 9x370 cuz my memory is not of the OCing type and I want to keept it close to stock speed, therfore I haven't tooled around with CPU VTT past 1.4. Also I haven't found too much of a difference in the CPU GTLREF. In some cases different settings seem to help a little but I haven't tested it thouroughly. I was hoping to reach a stable 3.4ghz but I get a little more peace of mind with the voltage under 1.4.

How high can you safely raise the vcore and still get at least a couple of years out of your processor? Is it more dependent on heat than raw voltage? If this is the case I guess I don't have to worry, I have a zalman cnps 9700nt that is performing marvelously! When trying to calibrate speedfan for temp I found that I couldn't do it. The guide states that you can't have CPU temps below room ambient, but the 9700nt, according to uguru and speedfan, had the temp at 12c with an ambient temp of 17c! I have been wondering if my temps have been correct because under load in prime I can't get it over 44c. also, my pwm temps hover around 60c max, and my cores are all running below 65c. Are my low temps just the result of keeping my home in the mid to low 60's? (it's cold here in Michigan so to me 60+ Fahrenheit feels warm!)

Anyways, just my thoughts. I have been following this forum and have found it very helpful. Thanks everyone!

[Supershanks]

Uguru CPU Temp is crap. do yourself a favour & get CoreTemp

[ScubaMob]

Yes, uguru IS poop when it comes to reading CPU temp. Coretemp confirms speed fans temps. Also, after some careful calibration it turns out my *real* idle and load temps are 25c and 46c respectively. Why do I need almost half a dozen programs to monitor my various systems accurately?! I'm pretty sure it's a conspiracy but I'm having a hard time coming up with a motive.... I'd better contact george bush so he can get homeland security working on this. While they probably wouldn't find an answer it would be the first time my tax dollars went towards something that might actually help me. Computer frustration... when will I stop being a noob?!

thanks, supershanks.

[Ace-a-Rue]

Abit late in the day but Happy New year people

Comparing memset Performance Level Gigabyte DQ6 v Abit IX38 Quad

I've just been playing & would appreciate your input Scope for Improvement?? It's something i found between the gigabyte Dq6 & the Abit. would be interested to see if ny results are repeatable & therefore gives hope that we might expect a better performance bios , should the Bios engineers stumble across it and tighten settings. Alternatively it may prove my results are
just statistical quirks.

cheers.

i think the bios engineers purposely set the timings to be loose on the first several releases...they don't want users to stumble coming out of the starting gate..you know, no post screen or can't clock very high.

i see the 7/8 PL on the DFI board when it is set to auto in the bios...on the Pro board i would see 7/8 PL...i miss my freaking Pro board...i am NOT a happy camper with this DFI board...it works but it drives me crazy with the multiple triple re-posts to get the settings to work...i can make a minor change, nothing to do with memory or mhz and it will shut down and re-start...when it does that, i am in it for the long haul to get it to show a post screen again!

[Supershanks]

Hi Ace,

i think the bios engineers purposely set the timings to be loose on the first several releases...they don't want users to stumble coming out of the starting gate..you know, no post screen or can't clock very high.

Good point Ace, especially when there's plenty else fot them to trip over. Probably a bit unfair , feeling a bit cynical this am.

Sorry to hear about your DFI struggle , but if anyone can whup that board into shape it's you , understand your frustration

luck

no problemo ScubaMob

[Supershanks]

New Bios Johnny bravo sent it See last item in BTABios Collection Looks good.

[jcool]

Just got back from skiing, felt like hell freezing over at -25C + chill factor
Anyways, gonna try this new bios asap and post results

[Supershanks]

Hi m8,

felt like hell freezing over at -25C + chill factor

good for ocing.
Wouldn't mind a couple of inches of snow so that i can get the tobogan out

Bios is ok it fixes temps as well, within 5C of coretemp now

look forward to see what you think:0

[Ace-a-Rue]

Hi Ace,
Good point Ace, especially when there's plenty else fot them to trip over. Probably a bit unfair , feeling a bit cynical this am.

Sorry to hear about your DFI struggle , but if anyone can whup that board into shape it's you , understand your frustration

luck

no problemo ScubaMob

thanks SS!

i think that i know what my problem was, but i need further verification from other DFI X38 posters, ...i think that i had a DEFECTIVE 9650...outward sign was core #0 was stuck (frozen) on 51C...if you ran it at stock speed and then ran wPrime 1024M bench the temp on core #0 would not budge...same way for idle conditions at stock speed, locked on to 51C.

i called Intel tech support to see if i could get an RMA and the guy wouldn't do it...he said DON'T worry about the temps...MAN, that torqued me off to hear him say that...he said if the processor gets to hot, it will signal the board to shutdown..so...no worry if you melt the silicon

i am going to type a letter to Intel headquarters to complain about the lack of support and not fulfilling their warranty.

any way, i was a day within my 30 day return policy for the place that i bought it at...the place is taking it back and they will send me a new one...i hope it isn't the same pack date...LOL

the only problem with this return was that i recycled the box and plastic..hehehe...so...they might charge me a 15% re-stocking fee...i hope they will feel pity on me and not charge me that fee....i will be eye balling my CC bill.

[redmodel]

Hi,

I just I've ordered a couple of these boards. I really wanted to get one ix38, and one DFI x38-t2r (so I can get to compare the two), but it wasn't in stock and I couldn't wait, and no 780i boards around for me to order either.. But that's alright, it'll be a while before I get a Nvidia chipset again after the whole Yorkfield with the 680i, and I have a feeling my old 680i board got flakey with quad cores at high voltages.

Anyhow, along with the boards, I ordered a few Ultra Mini Kaze 40mm fans to go over the NB and mosfets, along with Arctic MX-2 Thermal Paste.

My question, has anyone replaced whatever was inbetween the PWM and the heatsink with MX-2 or Ceramique? Was there a difference in temps?

Also, Can the Thermalright HR-05 SLI/IFX fit over the NB, and another on the SB? I noticed the NB seems to have a unique Hole placement, compared to other boards.

Looking forward to getting this board going, getting a new q6600 in too. You know, it was just my luck to get a L737B that had a vid of 1.325 Hopefully, the next one will be a bit better.

[Supershanks]

.i think that i had a DEFECTIVE 9650...outward sign was core #0 was stuck

Hope you've tracked it down then. my 1st E6600 i got last Jan was faulty, took a while to track down, rma on commando 1st , got new one still same instability, even at stock. So took e6600 for rma, got new one sweet as a nut. Talking to the tech support guy in the store & was only the 2nd person to get a faulty one. Their the equiivalent of newegg here in the uk. Scan

i am going to type a letter to Intel headquarters to complain about the lack of support and not fulfilling their warranty.

Hehe give them hell

take care

[jcool]

My question, has anyone replaced whatever was inbetween the PWM and the heatsink with MX-2 or Ceramique? Was there a difference in temps?

I wouldn't recommend using anything other than thermal pads on these mosfets, because they are really really easy to kill
Might not even be your fault, there only needs to be a minimal difference in height between these 5 chips and you either kill one of them if it stands off, or one goes up in smoke when it's lower than the others.
I wouldn't do it

Also, Can the Thermalright HR-05 SLI/IFX fit over the NB, and another on the SB? I noticed the NB seems to have a unique Hole placement, compared to other boards.

I don't think it fits.. those nb holes are really quite strange, w/o some modding you won't be able to fit any other cooler onto it atm - neither water nor air.

Looking forward to getting this board going, getting a new q6600 in too. You know, it was just my luck to get a L737B that had a vid of 1.325 Hopefully, the next one will be a bit better.

Ah, don't curse it just yet, I have a 737A with 1,225V and really having temp issues because it gets so f****** hot
It does run 3,6G@1,32V but I'm already at 65C core which is about the limit for stability. I tried one Q6600 with 1,325V Vid, and even though it needed 1,4V for 3,6G it only hit 55C or so - same cooling
I think I need to lap my CPU or something.. or go phase change

[GARY1951]

Need some advice please. I have narrowed down my buying decision for new system to this board or a Asus P5E. I have also decided to watercool my system. Any thoughts? Is the Abit board the way to go?? Any help would be appreciated. Thanks

[Ace-a-Rue]

Hope you've tracked it down then. my 1st E6600 i got last Jan was faulty, took a while to track down, rma on commando 1st , got new one still same instability, even at stock. So took e6600 for rma, got new one sweet as a nut. Talking to the tech support guy in the store & was only the 2nd person to get a faulty one. Their the equiivalent of newegg here in the uk. Scan

Hehe give them hell

take care

i think that i purchased from scan many years back...an asus AMD board that was only being distributed in the UK and europe....5 of us went in together to buy the board...i had a dickens of a time to complete the deal...i was talking via email with a sales person but i think he thought the deal was to complicated since it was an international order...i finally emailed the webmaster and he stepped in to help...got the deal done and i was very HAPPY!...so was four other guys around the USA...the boards came to me and i shipped them out to the other guys.

as you can see, i am TENACIOUS!...LOL

[andre X_X]

i tried the B08 bios and i cant boot my Q6700, it works great with my E6600 though... have to flash back to B06 to run Q6700

[Ace-a-Rue]

man, that's no good...BUT...i had similar experience with the IX38 and my QX9650...multi's wouldn't go above default multi with 12.B03...i had to retreat to 12.B02 to allow me to increase the multi above 9X

[ante_ante]

Seems like divider 1:25 doesn´t work to good in b08 either. Atleast not for me :/

[Supershanks]

@TENACIOUS! Not often we have tech that you don't We moved up to scotland in july, but prior to that was 15minutes away from sacn's sales counter.Was great for getting things quick & rmas Gave Clunk a lift up there a few times.

@andre X_X
That's wierd with the Q6700 & 12.08. Have you tried pulling the battery, letting dmi rebuild, to see if that makes any difference??


@Turbodream
No it's the same for us all. One day maybe

[ScubaMob]

Off topic...

I'm looking at the CPU VID in coretemp (q6600) and it shows 1.1625V. However, when it first opens it jumps back and forth from 1.325V to 1.1625V very fast for a couple seconds and than stabilizes at 1.1625V. which reading do you think it is? I didn't trust the reading with my overclock so I dialed it down to stock multiplier and FSB and it says the same thing.

Lower VID is better, generally speaking, right? 1.1625 VID sounds awfully low... is it a bull$hit reading? I also read that q6600's OC better with 8x multi rather than 9x multi, maybe that's why I can't get true stability @3.4ghz. I was just curious because I was thinking of RMAing the chip because it's cores show 11c difference at idle and it bothers me... it's either RMA it or do some lapping... or maybe it doesn't matter...

anyways, thanks all!

[farookh]

12.08 working good if only abit can fix the 1:1.25 in the final

Got up to 510x8 on a Q6600

[ScubaMob]

I must be doing something wrong...

I'm running 12.B06 with a q6600 @375x8 (supposedly 1.1625 VID), just tried to run prime with everything set at stock voltage except a little extra vcore... yeah, it locked immediatly.

Maybe I need to try some other BIOS? I can run at 3.33ghz with x9 multi but my vcore seems a little high at almost 1.4V

I'm trying to work up the clock slowly but I'm having no luck with x8. when you get the FSB near 400 what kind of VTT and MCH voltages should I be looking for?

[Supershanks]

8x multi rather than 9x multi,

Don't think so pps just like the higher fsb for better memory clocks.

Try
9x400 12HR Prime Stable as a base. Others have needed ICH 1.05V VOLTAGE to be 1.2V or above for stability. My vid is 1.2625v so you may need a little less vcore.

Got up to 510x8

Nice going

luck

[jcool]

Holy crap guys... I think my board's dying on me
I already told you about the strange sounds it's giving off even when in idle, but now it seems like the phase chips are definitely going to die or explode or whatever - just look at this:



Phases 1, 2 and 5 are getting way too hot after only a few seconds of prime, and that current of 162A is just ridiculous
Used to be like 94A@1,32V for 3,6Ghz. So if µguru is right, these PWMs are near death... Also Vcore is very unstable, I now have idle Vcore ranging from 1,35 to 1,39V and a Vdroop of 0,08-0,1V
Note that I didn't change anything at all with my hardware or cooling, still the same mips PWM freezer sitting tight on the mosfets.
I also never gave this board anything over 1,4V Vcore so I just don't see why it should crap out just like that
Will be my 1st Abit ever to go to RMA, hope they give me a decent one next time

[blind_ripper]

now thats hot , get a fan and more preseur on that sink !

testing my self: E6850 @ 3.8ghz 1.45vcore bios = 1.46vcore os idle 1.45vcore load .
phase chips get 43°C max 42A , this is with bios 12.B02

[dinos22]

digital phase i presume