**Official ABIT IP-35 Pro Review/Overclock/Guide Thread**

IP35 Pro, Q6600. I run at stock speed, but sometimes use the uGuru utility under Windows to adjust the FSB and Vcore to get 3 Ghz. It's 8+ hours Prime95 stable at that speed. But...

I'm seeing weird results when benchmarking with SiSoft Sandra. If I make the FSB and Vcore adjustments in the bios instead of using the Windows based utility, then Sandra sees the processsor as a 3.0 Ghz processor and my processor arithmetic scores are where they should be @ 55,193 ALU and 38,258 FPU. But, if I use uGuru from within Windows and use the exact same settings, Sandra sees the processor as 2.45 Ghz and the scores are 45,059 and 31,243 respectively.

Other benchmarks give me normal scores and ramp up with the processor speed when I use uGuru to overclock. CPUZ also sees the processor at the correct speed when OC'd through uGuru but Sandra doesn't. Does anyone else see this weird behavior with Sandra and using uGuru to OC from within Windows?

it makes you wonder about the effectiveness of sandra.

Quote:

---

Originally Posted by Ace-a-Rue

here is the latest Uguru (3.104): ftp://ftp.abit.com.tw/pub/download/u...s/uguru/3.104/

if you are using 3.104 then i would clear CMOS with the battery out for a few minutes...also push your power button and reset button to unload the capacitors while clearing CMOS.

BTW, what bios are you using on your board?...did you flash it with the "wb" (write boot block)...that area stows the voltages and NB memory strap settings...it is not normally flashed

if you are using the 11 official bios and clearing cmos does not take care of your issues, i would think about RMA'ing the board...but, try re-flashing and use the "wb" switch...follow my link in my signature to get the instructions on how to flash.

---

Uguru was 3.103, d.l 3.104 as I type. I did flash to the 11 official as per your instructions when I first got the board. I will try again when I take the board down to try a new hs that I want experiment with, but it needs a lap-job before I switch form the Ultra120. I am waiting for 2 new fans before I do the switch. I may do the bios again before then just to test. Thanks for the ideas.

Quote:

---

Originally Posted by Pntgrd

Uguru was 3.103, d.l 3.104 as I type. I did flash to the 11 official as per your instructions when I first got the board. I will try again when I take the board down to try a new hs that I want experiment with, but it needs a lap-job before I switch form the Ultra120. I am waiting for 2 new fans before I do the switch. I may do the bios again before then just to test. Thanks for the ideas.

---

which heatsink are you switching to?...i love my TRUE 120!

Quote:

---

Originally Posted by Ace-a-Rue

it makes you wonder about the effectiveness of sandra.

---

I'm not sure if it's Sandra or something on my end. I was hoping someone could verify.

Quote:

---

Originally Posted by Ace-a-Rue

which heatsink are you switching to?...i love my TRUE 120!

---

I am just going to try out a Gemini II. At $15 I wanted to see what effect it had on cooling the whole top of the board. If you thinnk about the design of it is just made for a Stacker 830. I can always go back to the Ultra if it doesn't work out but to tell you the truth, I am getting so frustrated with this Intel system I may pitch the whole thing and go back to AMD. I reflashed the bios again and same thing. Nothing above 3.5 is stable and I am not even sure what I HAD stable a couple of nights ago will be stable now. It took a restart at what WAS stable after setting the bios to get into Vista. I am half afraid to try any tests. It just is weird that if I try above 1.48 or 1.49 it justs reboots. I have been running my 1066 Ballistix at 1:1 5 5 5 15 so it cannot be that I would guess. PMW temps never get above the high 60's, cpu never above 61c on the highest core. I have spent a month reading up on overclocking this thing and 3 weeks playing with it. Something in the parts is a dud and I can't figure out what it is.

Quote:

---

Originally Posted by GameGuru

Also I noticed that you moved on the the DFI board, how you liking that?

:)

---

Either Im stupid or thiers something in my rig thats killing all these mobos cos the first DFI I got didnt last 1 day... Weirdest thing too that it boots with my E6600 but not my E6700:confused: , went to the store and let them test it out for themselves, they got the same results... the only difference between the 2 cpus is the E6600 was lapped and the 67 wasnt.... The e6700 was confirmed to be working, wierd shiet lol....

2nd board is running great, tested it with ln2 and I got higher results due to more available voltages... not gonna bother with the transpiper but from monitoring the chipset and PWMs temps, it should help and not just a gimmick as I originally thought.... Cooling the PWM heatsink dropped the temp for both the PWM and chipset by the same amount I think thats a good info for IP35 users since it doesnt have temp monitor for the chipset.... with no fan the chipset was 53c and pwm 51c, with a lil fan on just the pwm heatsink, temps were chipset 37c and 35c for the pwm....

Love to back using setfsb but this board does boot alot slower than the IP35p, if the extra voltage is not needed, then its not worth the extra $$ compared to the abit...

Folk, sandra is a nice program in general. But that odd stuff you talked about is simply they are behind with the prog a bit. p35 is an extremely new chipset, so dont expect alot of benchmark/monitor programs to fully support accurate data. Hell intel's download portion of their site still has yet to add P35 under >download&support>chipsetsoftware>...

Quote:

---

Originally Posted by luihed

Either Im stupid or thiers something in my rig thats killing all these mobos cos the first DFI I got didnt last 1 day... Weirdest thing too that it boots with my E6600 but not my E6700:confused: , went to the store and let them test it out for themselves, they got the same results... the only difference between the 2 cpus is the E6600 was lapped and the 67 wasnt.... The e6700 was confirmed to be working, wierd shiet lol....

2nd board is running great, tested it with ln2 and I got higher results due to more available voltages... not gonna bother with the transpiper but from monitoring the chipset and PWMs temps, it should help and not just a gimmick as I originally thought.... Cooling the PWM heatsink dropped the temp for both the PWM and chipset by the same amount I think thats a good info for IP35 users since it doesnt have temp monitor for the chipset.... with no fan the chipset was 53c and pwm 51c, with a lil fan on just the pwm heatsink, temps were chipset 37c and 35c for the pwm....

Love to back using setfsb but this board does boot alot slower than the IP35p, if the extra voltage is not needed, then its not worth the extra $$ compared to the abit...

---

wow man you've had more trouble with boards than me. and thats saying alot. i honestly don't know what your problem could be, i would look at trying a diffrent psu if you haven't already!!! I havent used water cooling since nforce 2 days..back when actual car rads where being used lol. IT has to be something with your block, or to much draw from the pump shorting something on a rail..who knows..thats nuts

good luck..

Quote:

---

Originally Posted by GameGuru

What are you smoking :confused:

---

urban weed

xfi is ok, its just the drivers. And to many issues with the cards, ive had nothing but trouble with my xfi since my a8n prem. I just got my HT omega striker in tonite, and it is crisp and pruned. Works fine, although the vista 64 driver is beta..still works good. EAX 2.0 though, so for the audiophile with games..they may be turned off. DTS to my onkyo sounds awesome though. for 80 bucks iam impressed, i went with the striker instead of the more pricey cla'ro model

anyway sorry to throw this off topic! back to ip35!

I was thinking the same about the psu but ive checked the rails ( with multimeter )and theyre all very stable... Power should be more than plenty cos my Decathlon 750w is only running the cpu, mobo, rams, 1 pci video, 2x raptors and thats it... my pump is AC and and the 3 fans for the rad runs on its own PSU... while benching any other fans used is geting power from my other rig.....

My setup is very lean and sits flat on my stereo stand so grounding is not an issue.... what I dont get is Ive been using the same HW since feb. and had no problem till P35... ive palyed with commando, P5B, 2x DFI 965-s also my first IP35 pro was trouble free for almost a couple of months then i the last 3 or 4 weeks I went through 7 mobos:confused: .... no major changes in my rig also during those killing times which makes it hard to troubleshoot.... only change zi did was getting an e6700 but it also died along with dead mobo #1:( ... best cpu Ive ever had too ( 567fsb and 4100mhz+ 32m stable on water ).... Maybe theres little elves that plays with my comp when Im sleeping lol....

BTW, none of the mobos died while running, theyll be fine then just boot anymore the nest time I use it:confused: postcodes I get are 8.7, C1, 9.9, 8.3 ... Ive seen these same issues with other people, I just went though all of them lol...

But this board I got right now is running strong.... **knock on wood**

Quote:

---

Originally Posted by strange|ife

urban weed

xfi is ok, its just the drivers. And to many issues with the cards, ive had nothing but trouble with my xfi since my a8n prem. I just got my HT omega striker in tonite, and it is crisp and pruned. Works fine, although the vista 64 driver is beta..still works good. EAX 2.0 though, so for the audiophile with games..they may be turned off. DTS to my onkyo sounds awesome though. for 80 bucks iam impressed, i went with the striker instead of the more pricey cla'ro model

anyway sorry to throw this off topic! back to ip35!

---

Are you having any trouble using the mic on the x-fi? I also have 64bit w/ 4gigs of ram and an x-fi, but I had to use my onboard sound for my mic and x-fi for my actual sound :rolleyes:. A LOT have been reporting the same issues. Supposedly the march 64bit drivers didn't have any problems at all, but then again there are obviously compatiblity problems with different programs because there would be no other reason to release new drivers. LMK if you have the same problem with the mic.

PS: Don't use windows to test out your mic. It works with windows. It's different programs that actually use the mic that have problems with it. More like crackling or clicking noise. Mine sounds like a helicopter :D

Quote:

---

Originally Posted by Pntgrd

I am just going to try out a Gemini II. At $15 I wanted to see what effect it had on cooling the whole top of the board. If you thinnk about the design of it is just made for a Stacker 830. I can always go back to the Ultra if it doesn't work out but to tell you the truth, I am getting so frustrated with this Intel system I may pitch the whole thing and go back to AMD. I reflashed the bios again and same thing. Nothing above 3.5 is stable and I am not even sure what I HAD stable a couple of nights ago will be stable now. It took a restart at what WAS stable after setting the bios to get into Vista. I am half afraid to try any tests. It just is weird that if I try above 1.48 or 1.49 it justs reboots. I have been running my 1066 Ballistix at 1:1 5 5 5 15 so it cannot be that I would guess. PMW temps never get above the high 60's, cpu never above 61c on the highest core. I have spent a month reading up on overclocking this thing and 3 weeks playing with it. Something in the parts is a dud and I can't figure out what it is.

---

a couple things that i would recommend:

do not use 1T....that can disrupt a lot of things when memory is being accessed at that speed....intel chipsets are all about high latencies...1T doesn't work very well on intel chipsets since MCH is in the NB...it might on AMD but that is because the MCH is located on the CPU die.

2nd, i heard via the grapevine that WD Raptor drives are subject to errors on this newest version of the ICH9(R) chipset(s)....if you can swap out the drive with another type other than a Raptor, you might just get a higher overclock.

3rd, try boosting your ICHIO 1.5v to 2.0v for a short time along with boosting your MCH to 1.7v for a short time to see if that steadies your overclock or allows you to go higher...if it does allow for a higher overclock or makes it very stable then you can back it down a notch at a time to find the sweet spot.

Quote:

---

Originally Posted by Ace-a-Rue

a couple things that i would recommend:

do not use 1T....that can disrupt a lot of things when memory is being accessed at that speed....intel chipsets are all about high latencies...1T doesn't work very well on intel chipsets since MCH is in the NB...it might on AMD but that is because the MCH is located on the CPU die.

2nd, i heard via the grapevine that WD Raptor drives are subject to errors on this newest version of the ICH9(R) chipset(s)....if you can swap out the drive with another type other than a Raptor, you might just get a higher overclock.

3rd, try boosting your ICHIO 1.5v to 2.0v for a short time along with boosting your MCH to 1.7v for a short time to see if that steadies your overclock or allows you to go higher...if it does allow for a higher overclock or makes it very stable then you can back it down a notch at a time to find the sweet spot.

---

Memory is set at 2T. At the moment I am using Vista which is on a Seagate drive, the Raptor is my XP install so that should rule that out. I had not played with the ICHIO voltages and MCH I have not gone above 1.54 (I think that is the step). I will give that a shot and see what happens. Let me ask you a question. I see the posts of some people having problems with tightening the cpu cooler too tight. I wonder, is it possible to have tightened the screws on the heatpipe assembly too tight. Could this possibley be causing my problems?

Quote:

---

Originally Posted by Ace-a-Rue

...2nd, i heard via the grapevine that WD Raptor drives are subject to errors on this newest version of the ICH9(R) chipset(s)....if you can swap out the drive with another type other than a Raptor, you might just get a higher overclock...

---

Really? :( That sucks! I recently bought a Raptor X when they were on sale at Best Buy.

it might be the sata 1 drives...the earlier version...the newer ones might be fine....i am not 100% on this...i might be totally off base with even making the statement.

Quote:

---

Originally Posted by Pntgrd

Memory is set at 2T. At the moment I am using Vista which is on a Seagate drive, the Raptor is my XP install so that should rule that out. I had not played with the ICHIO voltages and MCH I have not gone above 1.54 (I think that is the step). I will give that a shot and see what happens. Let me ask you a question. I see the posts of some people having problems with tightening the cpu cooler too tight. I wonder, is it possible to have tightened the screws on the heatpipe assembly too tight. Could this possibley be causing my problems?

---

most of the tightening problems are with water blocks or phase systems.

here is P95 inside the case, door closed at 3.8Ghz...actually still priming in the background while i make this post.

Hey Ace, could you or someone else if you have the time post a really quick list of programs/software that everyones using for benching please. More importantly what versions of the programs including Bios. Also if there is anything that comes to mind with a particular program could you shed some light on that as well? It seems that I read different issues with different versions of software. One last thing, I'm only talking for CPU, 3D is not needed. So for example,

1) uGuru - version
2) CPUZ - version
3) Super PI - Version
4) Core Temp - Version
5) Most stable BIOS

You guys get the idea. I know it may sound a little stupid but gosh this would help so much rather than sifting through thousands of posts.

Thanks guys :up:

Quote:

---

Originally Posted by GameGuru

1) uGuru - version = 3.104
2) CPUZ - version = 1.40
3) Super PI - Version = mod 1.5 XS
4) Core Temp - Version = 0.95.4
5) Most stable BIOS = you need to try them all

You guys get the idea. I know it may sound a little stupid but gosh this would help so much rather than sifting through thousands of posts.

Thanks guys :up:

---

answers inside your quote

Quote:

---

Originally Posted by Pntgrd

Let me ask you a question. I see the posts of some people having problems with tightening the cpu cooler too tight. I wonder, is it possible to have tightened the screws on the heatpipe assembly too tight. Could this possibley be causing my problems?

---

Somebody somewhere thoughout this planet of posts did have stability problems caused by overtightening their heatpipe.

Quote:

---

Originally Posted by AstroCreep

Really? :( That sucks! I recently bought a Raptor X when they were on sale at Best Buy.

---

There are a few people with problems and the 150Gb Raptors.

http://www.tomshardware.co.uk/wd-rap...view-2336.html

Personally I have two new 74Gb versions, both in Raid and single mode working fine for weeks. All with the IP35 Pro, Vista and a 8800GTX.

No problems here, my main drive is a 150gb Raptor.

Ace, that's a damn good clock at stock voltage. I can't believe you have a quad doing that while my E6850 won't go over 3500 without upping the juice.

What's your Super Pi 1M at that speed?

Quote:

---

Originally Posted by fornowagain

...Personally I have two new 74Gb versions, both in Raid...

---

Ditto, no problems.

When I got uguru systray-->about: EXE Version 3.0.2.3, but OCGuruOLE.dll's 3.1.0.4. Guess I better check the site. :D

damn, just did the heatpipe mod and added a 80mm fan on top of the pwm area, 20mm fan on both the pwm and NB heatsinks. Now my PWM temps are improving, but the NB runs about 44c idle and 50~51 when im playing games like MOH airborne... Would that be ok for the NB chip or it may dmg it? =S