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

[jcool]

Oh I just don't believe it. I just refuse to believe it.
Looks like my second IX38 just crapped out just like the first one. Ran 1 week folding, just stopped the program and all of a sudden got these buzzing sounds again
And sure enough, µguru showa 100A in idle (!) and 162A under load just like it did last time. Unstable Vcore, too. I feel like history repeating itself
Wasn't even on high voltages (1,33V CPU 1,4V MCH and so on) and the mobo is watercooled, after all. So wtf - can't this thing take even a week of continuous load? Have I had enough bad luck to get two boards from the same bad charge?
What else could it be? Maybe the PSU, but why now... I've had my Liberty 500 for over a year now and it ran fine with the old QuadGT.
I think the board is to blame.

[Humble]

humble - sorry to see your making coasters m8. Usually when there is a rom upgrade there is an upgrade of drivers to support the new rom. It's possible you need the updated driver, the last driver on abit's site is v1.17.19.02 D.2007/04/07 so either they've not got it up yet or it looks like 12.B08 is best bet for now, as you say.

My... I think I forgot that new driver option. But, I'm using JMB363 driver version exactly as you mention above since the first time.

But telling you the truth... My previous experience with AB9 QuadGT gives me big nerves to retry that 12.B09 or final BIOS. Back then, I had tried sooo many different attemps on that AB9 QuadGT board to persuade my JMB363 work properly, but all failed at that time:
- Update to newest BIOS version 15
- Download the newest JMB driver version 1.17.21 WHQL (have tried SCSI mode and native IDE mode)
- Use native M$ Windows built in IDE driver
- And even reinstal my Windows!
Everything was in vain at that moment. I then had no other choice, except to sell the board.

Thanks for your input anyway brother, but I think I'll best stick with this 12.B08 BIOS for now, as it give me minimum error. But it's a bit slow if compared to 12.B06 BIOS in Everest benchmark.
Btw, have you also find this issue on your system brother? IX38 BIOS 12.B09 with IDE DVDRW drives?

JMicron are not the best controller in the world, the sooner the motherboard industry takes that on board the better, I've lost count of the number of times these have causes people problems on a number of different platforms.

It may be worthwhile going to their (JMicron) site and seeing about updating the driver for the controller, this may help somewhat. Also just for clarification the abit BIOS engineers did not forget to change that stamp on that BIOS - the BIOS on their website is 12 the BIOS I sent out was basically the same but because it was not official they could not change the date stamp, in order to avoid two different BIOSes with the same date stamp appearing.

Thank you Brother Johnny for your information about the BIOS. It's all clear now. But still, I do hope that abit engineers make a more good, stable BIOSes soon without readding previous BIOS errors in the future.

About the J-Micron, I still can remember the time when Intel started to remove IDE drive support on their new chipset. A very brave and controversial step, as there were still so many IDE drives on the market and SATA drives hadn't become common at that time. Even nowadays, SATA interfaces is just widely used on HDD and still implemented very few on CD/DVD drives. SATA CD/DVD drives itself still also have so many bug compared to IDE CD/DVD drives.
The motherboard manufacturers then didn't have the other choice except to integrate this JMB363 on their board to support IDE drives which are still widely used at that time. Abit AB9 QuadGT was one of the board, and the new implemented JMB363 really gave me hard time at that time. My other friends also had the same problem with this board as I did.
But, the one to praise is Gigabyte. Their restamped J-Micron controller, as known as Gigabyte SATA2 chip, worked flawlessly on the very first P965 board, 965P-DS3P that I used! Not even single issue come up. Great job really!

Hello m8's, I'm new

Thats why I fitted an ide to sata adapter on my ide dvdrw so I can avoid the jmicron controller all together.

This thread is kinda depressing me I spent alot of money to get this new rig, the DFI X38 TR2 was in stock but I choose the abit over it, now I'm not sure I did the right choice...

Great idea there to use the adapter, brother!

Don't regret what you have decide. Abit IX38 is a great board but with more simple BIOS (not as complex as DFI; which keeps me away from using it) and it will show it's full capability after a few BIOS updates just like what happened to IP35 PRO. You just have to pray that moment will come soon.

Humble

Abit's IP35 Pro board works very well!

i noticed on DFI's LT X38 T2R that it had a problem with the Jmicron controller dropping the SATA DVD drive offline about 30-40% of the time if i had Intel's Raid enabled...it's the same DVD drive that i have in my Pro system right now working flawlessly on the Intel controller.

i would recommend getting the latest Jmicron driver to see if that helps your situation

Then I feel happy for you too brother. I think it's because of the maturity of IP35 PRO bios.

Right now, if I don't miss remember, the newest version of J-Micron JMB363 driver on it's official site is version 1.17.31 WHQL, but I don't sure it'll help me out of the problem if I'm still using 12.B09 BIOS.
Well, I think I'm happy enough with this 12.B08 now, and will stick with it for a moment till abit release a new, proven good BIOS. Thanks for your input too.

[Beastage]

Ok, I decided to have a go at 400x8, on stock voltages... why is it stable?

I running a Q6600 G0 cooled by ultra-120

I did it to see how long before it crashes... but its stable on full load for a few hours now.

Does that make sense?

On the topic of the JMicron driver , it was last updated on December 17th , 2007, so no new driver.

[wilson]

Hello im another new boy on here so HI ALL!!!
Just one question as i have just built my system about 5 days ago so any advice would be great.
I have raid0 on my two drives and am using intel`s matrix manager but not the J-Micron driver as everything works with only the intel installed. I even have my dvd drive in a sata port and so far have had no prob`s with it .
I have the new bios from abits site and am wondering if i need this J-Micron driver or not? I have raid enabled in my bios onchip SATA Device and in onboard PCI Device..
Do i make sence so far

[Johnny Bravo]

Oh I just don't believe it. I just refuse to believe it.
Looks like my second IX38 just crapped out just like the first one. Ran 1 week folding, just stopped the program and all of a sudden got these buzzing sounds again
And sure enough, µguru showa 100A in idle (!) and 162A under load just like it did last time. Unstable Vcore, too. I feel like history repeating itself
Wasn't even on high voltages (1,33V CPU 1,4V MCH and so on) and the mobo is watercooled, after all. So wtf - can't this thing take even a week of continuous load? Have I had enough bad luck to get two boards from the same bad charge?
What else could it be? Maybe the PSU, but why now... I've had my Liberty 500 for over a year now and it ran fine with the old QuadGT.
I think the board is to blame.

jcool have you tried running it outside the case those currents are WAY too big and this is more than coincidence!

[jcool]

How can it be related to the case, if it runs 2-3 weeks w/o problems and then, suddenly, it dies?
It's kinda dead now btw, I made the mistake of shutting it down. It won't start now, shutting itself of after a few seconds, then staying with postcode 90. Tried other gfx card, removing ram, clear cmos, remove battery but I guess it's done for.
I just don't get it... it can't be the PWM freezer, cause the freezer doesn't make contact with anything else apart from the mosfets. And if it did, why would it run for a few weeks then die without me even touching the case?
Another reason could be my PSU, maybe these digital PWMs are very sensitive to spikes. Against this argument stands the fact that I have used my trusty Enermax Liberty 500 for nearly a year with my old QuadGT already (which has essentially the same digital phase unit), also running 24/7 folding.

So I'm gonna give it one last try... will test my third board without the PWM freezer with fan attached and see what happens.
The only explanation I have is that the pulse chip is getting too warm without airflow and that it just burns out when running high load 24/7.
We'll see.

[Johnny Bravo]

No there's something more going on your current is messed up - I've had mine what 3-4 months now running on and off 1.6volts with a Q6600 and that is an idle currnet of 62A peaking at 120-130A fully loaded.

In effect I think you have a short developing somehow. This could happen by molten solder forming under the slave VRMs due to excessive heat, but this is unlikely as you are running a water cooling loop over that. However I think it would be in your best interest to run the next board with the stock cooler and a 60mm fan on it to assess if its the MIPS cooler or something else that is the culprit.

[Johnny Bravo]

you say you ran 1 week of folding, did you keep a check on the PWM temps during this ?

[jcool]

Yeah I did... temps were between 70 and 75C which is quite ok.
And like I said, I will try the next one w/o the PWM Freezer. If it still burns out it's either my CPU being a mobo killer () or it's the PSU...

[Johnny Bravo]

sounds like something is up for sure, hope it all works out in the end!

[Daidalos]

This board is great i could never pass 3.6 with the Ip35 because of the vdroop probably. Now i can do 3,8 prime stable!
Vcore is 1,52 in bios goes 1,49 at load. I am still testing the memory.

[jcool]

sounds like something is up for sure, hope it all works out in the end!

Thanks m8, will keep you guys posted..
For now, back to P965 QuadGT

[Supershanks]

And sure enough, µguru showa 100A in idle (!) and 162A under load just like it did last time. Unstable Vcore, too. I feel like history repeating itself

Wow like Johnny says's somethings seriously wrong there m8. What VTT Volts you on ??
I'm on 1.36v & it's currently ticking away at 30~32A Idle

hope it's not another RMA M8

Edit:
grood going Daidalos that's the good thing about this board

[zfactor]

i see some of you have had the dfi lp lt or ut models.. have you switched and not looked back i am looking into the dfi boards now.. i was sure about getting one and the more i read the more people seem to have reported issues.. but then there are issues with all mobo's lol.. they are never all going to be perfect.

this compared to the p35 pro ?? do you think the x38 really made a oc difference for you or was it something else.

are the bios chips soldered on these or are they chips that can be replaced?

[Ace-a-Rue]

Yeah I did... temps were between 70 and 75C which is quite ok.
And like I said, I will try the next one w/o the PWM Freezer. If it still burns out it's either my CPU being a mobo killer () or it's the PSU...

what spec cpu do you have (45nm?)...what is the VID for your particular processor?

[Ace-a-Rue]

i see some of you have had the dfi lp lt or ut models.. have you switched and not looked back i am looking into the dfi boards now.. i was sure about getting one and the more i read the more people seem to have reported issues.. but then there are issues with all mobo's lol.. they are never all going to be perfect.

this compared to the p35 pro ?? do you think the x38 really made a oc difference for you or was it something else.

are the bios chips soldered on these or are they chips that can be replaced?

the DFI X38 board is good...i have one..it's back in the closet..LOL..hehehe...actually, i like the IP35 Pro board...the DFI has a complicated bios for memory settings...you could run it in "auto" and never think about it, but, most people like to extract the most out of their boards....when you do that, it can be frustating

the big thing about the Pro board, it has a large Vdroop...the IX38 closed the gap on that issue...i think there is a small vdroop under load for the IX38.

[zfactor]

lol wanna sell it..

do you think the abit can get up to the same speeds as the dfi.. most are telling me the dfi can run a quad stable at near 500 24/7.. think this is true or that the abit can do the same.. again 24/7...

thanks for the info..

[Johnny Bravo]

what spec cpu do you have (45nm?)...what is the VID for your particular processor?

Looks to be Q6600 unless I got that wrong Ace

[Ace-a-Rue]

Looks to be Q6600 unless I got that wrong Ace

i am also curious about the VID for that processor...i have a VID of 0.9625v and it draws a huge amount of power when i raise the vcore to 1.45v or 1.48v...it sends my PWM temp (Pro board) into the red at 1.45v....that NEVER happened before!

i am thinking his VID is lower around my number which could have overloaded the PWM....could be wrong...it won't be the first time.

[Ace-a-Rue]

lol wanna sell it..

do you think the abit can get up to the same speeds as the dfi.. most are telling me the dfi can run a quad stable at near 500 24/7.. think this is true or that the abit can do the same.. again 24/7...

thanks for the info..

i replied to your PM.

i put my good Q6600 (G0) in the DFi board and i got the same out of it that i got with the Pro board...the only real difference between the two; there is no Vdroop on the DFI board, besides the PCIe configuration....so...i didn't have to overvolt the DFI board....i still had to contend with the heat produced by the vcore...i used a TRUE heatsink.

[AMDfreak]

No there's something more going on your current is messed up - I've had mine what 3-4 months now running on and off 1.6volts with a Q6600 and that is an idle currnet of 62A peaking at 120-130A fully loaded.

In effect I think you have a short developing somehow. This could happen by molten solder forming under the slave VRMs due to excessive heat, but this is unlikely as you are running a water cooling loop over that. However I think it would be in your best interest to run the next board with the stock cooler and a 60mm fan on it to assess if its the MIPS cooler or something else that is the culprit.

In case it helps, I had this problem with my first IX38. I had the stock cooler with a 92mm fan directed on it. Without the fan, my PWM temps got to around 120C, with the 92 temps stayed under 90C. My CPU is a Q6600 G0 that I had vcore set as high as 1.6. The VID for my CPU is 1.2875.

I've got a 3rd IX38 on the way, so I'll let you guys know what's up.

[wilson]

Oc advice needed please.
Ok guys im oc`ing my Q6600 and am messing about with my multi and so far i am at 400 x 8 with the following volts
cpu 1.325v
cpu vtt 1.41v
mch 1.37v
ichio 1.50v
ich 1.11v
I am also running my 4gb of ram in 1.1 so it makes it 800Mhz.Do they look ok for volt`s ect or do i need to change anything b4 i start to bench as my cores are 38/38/39/41 on idle..And if i was to mount 2 x 120mm fans on my cpu cooler should i put them both blowing towards each other or not.
Cheers Paul.

[jcool]

what spec cpu do you have (45nm?)...what is the VID for your particular processor?

It's a Q6600 G0, the evil type - Vid 1.225 (means it gets very hot and draws lots of A). "Normal" full load operation at 3,6Ghz and 1,32V put 107A through the PWMs (was the same on P965 QuadGT). Only my X3210 B3 Xeon was worse, I saw that one sucking up to 125A@3,2Ghz

BTW, gotta love my Win XP x64 - just slapped in an IP35 Pro that happened to lie around (being a system builder does have its benefits) and it bootet straight into windows, including raid and all, just like it did when I changed from P965 to X38

Supershanks: Was running a VTT of 1,36, like I said it's NOT the voltages being too high.
I think we have narrowed down the problem to either the mips PWM freezer or my liberty 500.

@AMDfreak: U also on ur 3rd board? Damn this sucks... you don't happen to have an Enermax PSU, by any chance?

OT: Can anyone tell me which bios to use for the IP35 pro? Since I'm gonna have the board for at least a week why not test it out a bit.
But I'll have to put a fan over NB and PWM, currently running stock idle, CPUtemp 25, System temp 30, PWM 60 - told you my CPU was evil

[AMDfreak]

@AMDfreak: U also on ur 3rd board? Damn this sucks... you don't happen to have an Enermax PSU, by any chance?

Nope, PCP&C Silencer 750. Single 12v rail with 60A. My replacement for the 1st board was DOA. On power-up it would just do this code 90 to 99 to power off and back on again in an infinite loop. I did get it to post once and flashed to b09, but no change.

Edit: Are there any Irishmen around that can send me and jcool some good luck?

[jcool]

IP35 pro sucks for me - been running 400x9 and got a bluescreen after 2 mins of prime, PWM reached 95C *g*
Anyway why the hell can't one adjust the VTT properly?!
Now running 375x9, droop is kinda ridiculous... 1,28--->1,23 load... I gotta get my X38 back