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

[18 is # 1]

I can't get past 425 FSB.

[Ace-a-Rue]

O.K....got rid of the error permitting 2x2 Ballistix and put in OCZ Titanium CL3 6400. These are :

http://www.ocztechnology.com/product...l3_edition-eol

made for EPP and can show T1 timmings.

Now here is my question....Bios shows T1 and Memset shows T1 but CPUZ shows T2. Which is correct? I've read reviews that say non-SLI chip mobos can run T1 with these.

Although I've got some tight timings, I still can't get stable past 3.8Ghz (yes I tried 5-5-5-15 T2). I was hoping the RAM was holding me back.

forget about going past 3.8 ghz Prime 95 stable; not unless you have a Vapo unit.

[Ace-a-Rue]

I was able to get 476fsb bench stable for a q6600 but that was the absolute wall for that cpu.

it is just about the same for my QX9650 on a Gigabyte X48T-DQ6 (DDR3) and the same for the IP35 Pro...

[redrumy3]

You won't be able to push 500 fsb with a 65nm quad. I believe the Q6600's max out on FSB in the 460-480 range on this board. I'm not sure if that holds true for other boards as well.

The great thing about 400x9 though is that you can leave EIST enabled and it will remain stable, so you can let the system idle @ 2.4 ghz (400x6) and save a little power and reduce the vcore that the chip is seeing all the time.

ahh well that stinks, thanks for the info

[JamesBong420]

why would you want to switch to this board when you are doing so good with the DFI?...that is a hard one to figure out...i mean, 570 FSB is nothing to sneeze at!

the dfi is my hommies. it was in his mobo for a week or so. well i guess its new mobo time. the media center needs a better mobo anywho. but still any of the bios's make a big difference with the fsb on 45nm?

Which DFI was that?

DFI p35 DK - friends mobo as stated above.

[AstroCreep]

the dfi is my hommies. it was in his mobo for a week or so. well i guess its new mobo time. the media center needs a better mobo anywho. but still any of the bios's make a big difference with the fsb on 45nm?



DFI p35 DK - friends mobo as stated above.

Did your friend's DFI make a weird, squealing noise? I had a 965 for a week but I had to return it because it was so loud and annoying...

[18 is # 1]

Is FSB limited by the CPU, mobo, RAM, or a combination? I've changed the RAM, think I have a decent mobo, is it my E8400 limiting me to 440. 475 won't even boot XP.

[SetteUltras]

vcore= 1.45 (actual)...(be ready to give it more because of Vdroop under load)
MCH= 1.48 (might require higher)
CPU VTT= 1.30
ICHIO1.5= 1.70 (maybe, 1.80)
ICH1.05= 1.05 (or 1.09)
DDR2= 2.0 to 2.1
DDR2 REF= +2%
CPU GTLREF= 67-69% (for 65nm CPU's) (45NM CPU's go lower)

Thank you ace. I cannot keep it stable even with 1.5V. On load (priming) goes down to 1,4-1.41V.
I am really dissapointed with this board. Vdrop is too much and i am on air so i cannot try over 1.51V.
So far the only prime stable point i have is 8x420 @1.5V 1.41 on load)
Is vdrop getting less with bios 16 or is it the same? I usr bios 14 now.
I am really thinking going for a m/b having voltage calibration in bios menu....

[Imitation]

Is FSB limited by the CPU, mobo, RAM, or a combination? I've changed the RAM, think I have a decent mobo, is it my E8400 limiting me to 440. 475 won't even boot XP.

Loosen up your timings to 5-5-5-18 running 1:1 and see how high you can go if you haven't done that already.

[18 is # 1]

Loosen up your timings to 5-5-5-18 running 1:1 and see how high you can go if you haven't done that already.

Thanks, but I've tried all kinds of timings and voltages and dividers .

[loonym]

Man that would be unusual to see an e8400 hit the wall at 440

[Zinthar]

Thanks, but I've tried all kinds of timings and voltages and dividers .

Have you:

1) Disabled EIST and C1E?

2) Set Reference Memory Voltage to +2%?

3) Tried turning the multiplier down to 6 and upping the FSB to test for the wall?

[Ace-a-Rue]

Thank you ace. I cannot keep it stable even with 1.5V. On load (priming) goes down to 1,4-1.41V.
I am really dissapointed with this board. Vdrop is too much and i am on air so i cannot try over 1.51V.
So far the only prime stable point i have is 8x420 @1.5V 1.41 on load)
Is vdrop getting less with bios 16 or is it the same? I usr bios 14 now.
I am really thinking going for a m/b having voltage calibration in bios menu....

have you done the heatpipe mod?

[Zinthar]

Thank you ace. I cannot keep it stable even with 1.5V. On load (priming) goes down to 1,4-1.41V.
I am really dissapointed with this board. Vdrop is too much and i am on air so i cannot try over 1.51V.
So far the only prime stable point i have is 8x420 @1.5V 1.41 on load)
Is vdrop getting less with bios 16 or is it the same? I usr bios 14 now.
I am really thinking going for a m/b having voltage calibration in bios menu....

I also own a Q6600 G0 (VID: 1.2875v) and was able to get it stable at 400x9 and 450x8 by giving it 1.5175v in BIOS (which equates to about 1.43v under full load).

My vdroop is awful, and wasn't changed by going from BIOS 14 to 16 (nor was my overclocking potential or stability, from what I can tell).

I feel comfortable feeding 1.5175v in BIOS because HWMonitor tells me that the chip itself is never getting more than 1.48v maximum, and that occurs very rarely because I have EIST & C1E enabled (which works if you choose the 400x9 overclock, though likely not at 450 mhz FSB) -- so most of the time my chip is just getting 400x6 and 1.38v.

So if I were you, I'd turn C1E & EIST on, set your OC to 400x9, and up the voltage in the BIOS. If your small fft load temps start eclipsing 71 degrees, your cooling is the limiting factor. I had to lap my Q6600 and tuniq tower in order to keep my prime95 load temps under 70.

I have some complaints about this board, but Q6600 G0 overclocking potential (and stability at high OC's) is certainly not one of them.

What's the VID on your Q6600?

[18 is # 1]

Have you:

1) Disabled EIST and C1E? Yes
2) Set Reference Memory Voltage to +2%? Yes

3) Tried turning the multiplier down to 6 and upping the FSB to test for the wall?

Yes

Bought some Corsair Dominator TWIN2X2048-6400C3DF CL3-4-3-9

http://www.newegg.com/Product/Produc...82E16820145165

I figure if I can't go over 3.8Ghz, I'll have the fastest 3.8 there is!

Dam Ram Tootin'!

[Zinthar]

To update on the IP35 Pro's awful high frequency whining, I was able to minimize the problem by changing my Vista power settings (high performance made the CPU run at 100% all the time).

So I've isolated the problem to occuring only under high loads. I have a feeling that a different motherboard (or even another IP35 Pro) may not have this issue, but I've seen a million similar complaints from people with laptops/desktops of various builds, etc. so I guess it's common.

What is more disturbing to me is the fact that my parents' IP35 Pro doesn't have the same EMI problems that mine does (audible via the onboard audio). So I guess I can either keep the board and buy a sound card and live with the buzzing under high CPU load... or go with a new board altogether (maybe a DFI Lan Party) and hope the problem was just my IP35 Pro.

Anyone else out there hearing this buzzing under load?

[SetteUltras]

have you done the heatpipe mod?

Yes done the mod first day i bought the board. Actually with 1.51+ vcore i expect high temps but if the was no vdroop it would be stable at 1.43V which temps would be better overall. So mainly vdroop is my problem and cannot change unless h/w mod the board which i am not capable of doing!

@Zinthar; Mine is vid 1.2625V
I could do what you say but then i would have low memory b/w plus high latencies. My goal ofr 24/7 is the following

[Ace-a-Rue]

To update on the IP35 Pro's awful high frequency whining, I was able to minimize the problem by changing my Vista power settings (high performance made the CPU run at 100% all the time).

So I've isolated the problem to occuring only under high loads. I have a feeling that a different motherboard (or even another IP35 Pro) may not have this issue, but I've seen a million similar complaints from people with laptops/desktops of various builds, etc. so I guess it's common.

What is more disturbing to me is the fact that my parents' IP35 Pro doesn't have the same EMI problems that mine does (audible via the onboard audio). So I guess I can either keep the board and buy a sound card and live with the buzzing under high CPU load... or go with a new board altogether (maybe a DFI Lan Party) and hope the problem was just my IP35 Pro.

Anyone else out there hearing this buzzing under load?

occasionally a board shows up with that noise...it is generally the choke coils were not filled with enough silcone or has pockets without silcone...the silcone is used to keep the coils from vibrating under heavy load on the cpu...i would say to RMA the board back to the retailer if you can.

[Ace-a-Rue]

Yes done the mod first day i bought the board. Actually with 1.51+ vcore i expect high temps but if the was no vdroop it would be stable at 1.43V which temps would be better overall. So mainly vdroop is my problem and cannot change unless h/w mod the board which i am not capable of doing!

many boards have Vdroop...i have it on my gigabyte X48 (DDR3) board...it is not a lot but it is there...my IP35 Pro has about 0.04v vdroop...don't get wrapped around the axle with what the bios vcore is set to...or...what Uguru says is set...your main concern; what is actual idle vcore and then what is the droop when you load the cores to 100%...that is the "true" Vdroop.

[Imitation]

many boards have Vdroop...i have it on my gigabyte X48 (DDR3) board...it is not a lot but it is there...my IP35 Pro has about 0.04v vdroop...don't get wrapped around the axle with what the bios vcore is set to...or...what Uguru says is set...your main concern; what is actual idle vcore and then what is the droop when you load the cores to 100%...that is the "true" Vdroop.

Yeah my droop is about the same here with a new 45nm quad. It shows about 1.23 idle and 1.2 to 1.19 loaded with OCCT. That isn't much vdroop at all, my old asus P5N-E SLI board was like .07-.09 with only a dual core! I wasn't a huge fan of that thing, glad i sold it off and got this board

[18 is # 1]

many boards have Vdroop...i have it on my gigabyte X48 (DDR3) board...it is not a lot but it is there...my IP35 Pro has about 0.04v vdroop...don't get wrapped around the axle with what the bios vcore is set to...or...what Uguru says is set...your main concern; what is actual idle vcore and then what is the droop when you load the cores to 100%...that is the "true" Vdroop.

So if we want to protect the processor from too much voltage, do we look at CPU-Z or the settings in bios/uGuru?

[Zinthar]

occasionally a board shows up with that noise...it is generally the choke coils were not filled with enough silcone or has pockets without silcone...the silcone is used to keep the coils from vibrating under heavy load on the cpu...i would say to RMA the board back to the retailer if you can.

Actually it turns out that I killed my IP35 Pro...

I hadn't noticed the vibration before, and that's because it wasn't there... I overloaded the 4-phase PWM on the IP35 Pro in my attempts to test the stability of my 400x9 overclock. It turns out that neither the IP35 Pro nor the IX38 are good overclockers for massive voltages from 65nm quad cores, and the PWM can't handle the stress involved in most cases.

Unfortunately, it has been over 30 days since I bought the board from Newegg, so a replacement/refund from them is not an option, and I don't trust that I wouldn't just as easily overload the PWM of the replacement either... so I'm going to buy a replacement board and RMA this Abit IP35 Pro and sell it to a friend who wants to build a cheap E8400 rig with it.

It's a shame because the IX38 has a spectacular MIR right now on newegg, but I think I'm just going to get the more expensive DFI P35 T2R and deal with it now (some users have written about how their 3.6 Ghz Q6600 overclocks have killed the 5-phase PWM in the IX38 as well) -- it includes 8-phase digiital PWM, which I think is necessary to keep a Q6600 stable at 3.6 Ghz for the long haul.

I love everything else about the IP35 Pro, but I should have been more attentive to its ever-increasing vDroop as a sign of mine dying (my load vDroop is now 0.09-0.1v). Oh well... I would recommend it unquestionably for anyone looking to OC any Core 2 Duo.

[loonym]

Really sucks about the board. I think a broad qualifying statement about the board's ability to handle 65nm quads is grossly inaccurate. You fried the mosfets, major suckage but it sounds like you may have had a flawed board from the start. I found immediately that that area of the board in particular needs special cooling attention. Contact was poor and I consider the pad useless. My ip35-pro has run a q6600@3600 100% load nonstop crunching WCG since I set it up and I have no reason to believe it won't continue to do so for a good long time. Vdroop is a fact, I have it but it's not out of the range of any other board I run. I'm quite satisfied and may add another quad cruncher built around ip35-pro simply because of the simplicity and stability.

[Zinthar]

Really sucks about the board. I think a broad qualifying statement about the board's ability to handle 65nm quads is grossly inaccurate. You fried the mosfets, major suckage but it sounds like you may have had a flawed board from the start. I found immediately that that area of the board in particular needs special cooling attention. Contact was poor and I consider the pad useless. My ip35-pro has run a q6600@3600 100% load nonstop crunching WCG since I set it up and I have no reason to believe it won't continue to do so for a good long time. Vdroop is a fact, I have it but it's not out of the range of any other board I run. I'm quite satisfied and may add another quad cruncher built around ip35-pro simply because of the simplicity and stability.

I don't mean to make a broad statement, but what I mean to say is that there's evidence from a few users who have had the PWM hissing/death from multiple IP35 Pro/IX38 boards -- hardly what anyone would consider proof or even a reasonable study, but enough to make me consider that if I want to build a rig that will last 3 years with a strong 24/7 overclock of my Q6600, I better go with something that has a 6 or 8-phase PWM, preferably digital.

I'm fine with fairly large VDroop, but what has started happening with my IP35 Pro (and I still have it running as I type this, since I'm waiting for my new board to arrive before disassembling it) is that the VDroop under full load has been getting worse at the same frequencies. I originally had set my Q6600 to 1.5175 in BIOS, which was coming out as about 1.45-1.46v under Prime 95 small fft load. After all of the stress testing I did, by the end that same 1.5175v was reading as 1.42-1.43v. Same BIOS the whole time (16B04). The hissing has all gotten worse, and when I put my ear up to the board I can hear a lower frequency hum (note that I'm now running at stock speeds/volts again, but the hissing is still audible at idle now). I found an IX38 thread on a forum where a few quad core users were experiencing the same phenomena and reported that Abit technical support suspected the PWM's had fried.

I wondered why I hadn't heard the hissing sound the first week I had the board when I had my case open and it should have been readily audible to me -- I can only suspect that I didn't hear it then because the PWM's were fine then.

I would have confidence that I could do a re-build of my IP35 Pro and it would probably be fine, but I think I stand a better chance with DFI's LT P35 T2R and the 8-phase digital PWM. It'll be my first DFI board. Now I hope I don't kick myself for having skipped getting the IX38 or Asus Maximus Formula and their PCI-E 2.0 slots...

[JuliM]

Hi,
Has somebody else get some changes in Everest memory benches since updating to beta 10 BIOS? When I compare my results to SetteUltras for example, my writing bandwidth is always higher than the reading bandwidth (before the update is was the same as SetteUltras (Read>Write)).