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

[scorpion]

http://http://www.anandtech.com/trad...spx?i=3250&p=2

[Ace-a-Rue]

it won't connect.

[scorpion]

http://www.anandtech.com/tradeshows/...spx?i=3250&p=2

[scorpion]

looks like we have new ip pro on the way!!!!!!

[golfjam]

I just got one of these boards and need help. The temps are a bit crazy under load, speedfan shows 58, realtemp 50, core temp 51, everest 65. Which one is reliable?? I always trusted core temp but it normally shows higher than speedfan. I wouldn't bother but I changed my hsf from a Zalman 9700 to thermalright 120 x, so I'm not sure of what's going on here. Plus, is there a fix to run memory at 1:1.25 w/ 4g?? If I don't run 1:1, prime fails in 30 sec. I use bios beta 16 10. I know it's been answered before but it's hard to filter in 227 pages. Thanks for the help.

[scorpion]

http://www.anandtech.com/printarticle.aspx?i=3250

[scorpion]

sorry about the first link!!!

[Ace-a-Rue]

I just got one of these boards and need help. The temps are a bit crazy under load, speedfan shows 58, realtemp 50, core temp 51, everest 65. Which one is reliable?? I always trusted core temp but it normally shows higher than speedfan. I wouldn't bother but I changed my hsf from a Zalman 9700 to thermalright 120 x, so I'm not sure of what's going on here. Plus, is there a fix to run memory at 1:1.25 w/ 4g?? If I don't run 1:1, prime fails in 30 sec. I use bios beta 16 10. I know it's been answered before but it's hard to filter in 227 pages. Thanks for the help.

on my rig, i have about 10C split between real temp2.2 and coretemp since coretemp is using TjMax of 105C and realtemp uses 95C Tjmax.

realtemp is almost the same as Uguru temp readout so i am bending in that direction.

as to P95 failure after 30 seconds in other than 1:1, that generally has to do with insufficient voltage for MCH, ICHIO and Vdimm....i would increase MCH first followed by ICHIO and then Vdimm...the 1:1.25 is a tighter latency (1066 memory strap) so more voltage is needed to provide the push through power for the data to get back and forth between L2, CPU FSB, MCH hub and ICHIO hub....one other voltage to be concerned with, although far less, is CPU VTT...i would set that in the 1.3X, low end range.

also, don't forget about CPU GTLREF...a Q6600 needs in the range of 67-69% but experiment on the lower end too since bios updates might affect that point of stability.

[Ace-a-Rue]

http://www.anandtech.com/printarticle.aspx?i=3250

can't wait to jump on abit's IX48 max board...i hope they release it along with tier 1 board makers.

[scorpion]

yea, maybe we will see it in april!!!

[golfjam]

on my rig, i have about 10C split between real temp2.2 and coretemp since coretemp is using TjMax of 105C and realtemp uses 95C Tjmax.

realtemp is almost the same as Uguru temp readout so i am bending in that direction.

as to P95 failure after 30 seconds in other than 1:1, that generally has to do with insufficient voltage for MCH, ICHIO and Vdimm....i would increase MCH first followed by ICHIO and then Vdimm...the 1:1.25 is a tighter latency (1066 memory strap) so more voltage is needed to provide the push through power for the data to get back and forth between L2, CPU FSB, MCH hub and ICHIO hub....one other voltage to be concerned with, although far less, is CPU VTT...i would set that in the 1.3X, low end range.

also, don't forget about CPU GTLREF...a Q6600 needs in the range of 67-69% but experiment on the lower end too since bios updates might affect that point of stability.

I'll try and let you know. Man, I love this forum!!

[tenax]

if not posted yet, i'm told that a new official 16 bios is up at the abit ftp. have only downloaded it at this point..i'll wait for someone else to install and report on it

ftp://ftp.abit.com.tw/pub/download/bios/ip35/

edit- my bad..as i suspected...should have confirmed first, it's for the ip35..not the pro..hopefully, we're getting close though

[Ace-a-Rue]

if not posted yet, i'm told that a new official 16 bios is up at the abit ftp. have only downloaded it at this point..i'll wait for someone else to install and report on it

ftp://ftp.abit.com.tw/pub/download/bios/ip35/

edit- my bad..as i suspected...should have confirmed first, it's for the ip35..not the pro..hopefully, we're getting close though

that is the regular IP35.

go here for the latest bios for the IP35 Pro: ftp://ftp.abit.com.tw/pub/download/bios/ip35pro/

[rentay]

Hi all!

Just got my system up and running! All I have to say is wow! This Abit IP35 Pro motherboard is awesome.

My specs are as follows:
Intel Q6600 SLACR (G0)
Abit IP35 Pro (running Bios 14)
Corsair Dominator 8500C5 (5-5-5-15 2.2v)
Watercooled setup
WD Raptor X 10000rpm
Pioneer DVR-215 Dual Layer burner

I just finished overclocking the cpu using Prime 95 torture testing all 4 cores (I used Prime 95 v.256).

To monitor temps I used both core temp and real temp, but from reading some responses in this thread real temp is the more accurate to go by. On water, my cpu maxed 46 degrees celcius on cores 1 & 4 and 42 degrees celcius on cores 2 & 3.

Prime95 was run on torture test set to "blend" option for 6 hours and all is stable!

My overclock is as follows:

FSB = 400
Multi = 9
DRAM Speed = 1:1.25 (DDR2-960)
PCI Express Clock = 101Mhz

Voltages:
Vcore = 1.4675
DDR 2 = 2.20v
MCH = 1.29
and the remaining voltages standard

Under Ram, I manually selected options as follows:
CAS Latency = 5
RAS to CAS Delay = 5
RAS precharge = 5
Precharge delay = 15
and remaining set to auto

Any questions please feel free to ask! This thread has been an absolute dream! Thank you!

Regards

[golfjam]

on my rig, i have about 10C split between real temp2.2 and coretemp since coretemp is using TjMax of 105C and realtemp uses 95C Tjmax.

realtemp is almost the same as Uguru temp readout so i am bending in that direction.

as to P95 failure after 30 seconds in other than 1:1, that generally has to do with insufficient voltage for MCH, ICHIO and Vdimm....i would increase MCH first followed by ICHIO and then Vdimm...the 1:1.25 is a tighter latency (1066 memory strap) so more voltage is needed to provide the push through power for the data to get back and forth between L2, CPU FSB, MCH hub and ICHIO hub....one other voltage to be concerned with, although far less, is CPU VTT...i would set that in the 1.3X, low end range.

also, don't forget about CPU GTLREF...a Q6600 needs in the range of 67-69% but experiment on the lower end too since bios updates might affect that point of stability.

It's working, I've increased the CPU GTLREF to 69% and it's all rock solid. Thanks again for the help, I truly appreciated.

[chawks2]

..CPU GTLREF to 69% and it's all rock solid.

It's very bizarre to me the GTLREF value is so inconsistent. Mine is @ 64% and rock solid, Ace, last time we spoke, mentioned his was @ 61%. Why is there such a discrepency on this value?

[golfjam]

It's very bizarre to me the GTLREF value is so inconsistent. Mine is @ 64% and rock solid, Ace, last time we spoke, mentioned his was @ 61%. Why is there such a discrepency on this value?

Since you've mentioned, I decide to lower to 63%. Funny, Prime95 didn't stop and still running.

[Ace-a-Rue]

It's very bizarre to me the GTLREF value is so inconsistent. Mine is @ 64% and rock solid, Ace, last time we spoke, mentioned his was @ 61%. Why is there such a discrepency on this value?

since then, i increased (tweaked) my penryn, QX9650, to 63%.

when i was intially reading the tech documents, i read that the wolfdale cpu's should work with lower CPU GTLREF values of 63%...i took that to mean it applied to 45nm cpu's and found it worked good on my 9650...

as you know, every cpu is different as is motherboards for the same model, so it behooves anyone to experiment up and down from that value.

[chawks2]

..as you know, every cpu is different as is motherboards for the same model, so it behooves anyone to experiment up and down from that value.

Do you really believe the CPU to be the culprit, especially from the same batch? I would think mobo's would be the PITAs.

Anyways, I do agree with Ace on experimenting. Took me a long while to find my sweet spot, although it may have been due to my old RAM setup w/crossed vendors @ 8G.

[kadozer]

Just wanna say thanks to all that help out in this awesome thread. I finally did the bolt mod and wanted to post my idle temps with a Q6600 @ 3.4ghz. PWM dropped 9-10C idle and 7-8C load after the mod.

[wilson]

Hi guys is anybody using the e8400 in this mobo as im gonna put one in and if so is bios 12 still the one to use or should i use one of the 16 betas??
Cheers Paul

[BassKozz]

Any questions please feel free to ask!

What's your PWM temps like? (idle/load)
Screenshots?

Just wanna say thanks to all that help out in this awesome thread. I finally did the bolt mod and wanted to post my idle temps with a Q6600 @ 3.4ghz. PWM dropped 9-10C idle and 7-8C load after the mod.

What is your LOAD PWM temp?

My Idle PWM temps are 33~34c @ 400x9 & 31~32 @ 370x9...
My LOAD PWM Temp at 400x9 is between 67~70c , which I didn't like so I've since throttled back to 370x9 which gets my PWM LOAD temp down to 54~56c which I am more comfortable with.

I know I could probably use a fan to get the PWM down to my comfort zone @ 400x9, but I don't like the idea of rigging up a fan that close to my MoBo's circuits (I know I am a :banana::banana::banana::banana::banana:, but I don't want to hurt my pretty little IP35-Pro )

[kadozer]

Load Temps doing Prime95 'blend' right now as I type:
Vista x64, 4gb ram

After 25min EDIT: Increased my intake fan (blowing over motherboard) to 2400rpm wow nice PWM drop of 6-7C off load temp.


Just wanted to also add for anyone interested doing the PWM bolt mod I find that the fiber washers with metal/rubber combo work the best.
I've tried regular rubber washers stacked in 3s as well as the plastic bolt cover designs found at Home Depot and both weren't as effective with my PWM temp.

[lefy]

I did the bolt mod and just got the rig running again. How are my temps? PWM is 67C under load... no idea what it was before the mod

Prime95 was crashing after about an hour at 3.7GHz & I had no fan on the PWM or NB. They were scorching when I touched em. Now I've got a 12" floor fan blowing right on 'em and that seemed to be the trick. They barely feel warm to the touch now. Guess my NB or PWM temps were my limiting factor - I need to get some permanent fans setup.


oh, cpu idle was about 36C right after stopping a 3 hour prime95 run. & vcore is 1.4050v in BIOS

[Ace-a-Rue]

I know I could probably use a fan to get the PWM down to my comfort zone @ 400x9, but I don't like the idea of rigging up a fan that close to my MoBo's circuits (I know I am a :banana::banana::banana::banana::banana:, but I don't want to hurt my pretty little IP35-Pro )