Asus 680i - 450+ FSB Stable on Quad Core Possible With GTL Ref Mod. - 56k Warning.

the 680i isnt a very strong board when it comes to Quad Core overclocking, we all know that...
but what it does have is a great ability to tweak ram, and great SLI options.
speed for speed, it can use alittle work, but with some good tweaking, it can be just as good, or better then other boards out there.
except one thing always plagues it...
its Quad Core overclocking, is just kinda pitiful.
333-350 is the typical stable range, while you can get 6 x 450 to work on a unmodded board... if you only use the 6 multiplier. 7, 8, 9 and higher... become near impossible for some reason.

well, I saw a few weeks back that Shamino released a great little guide to modding the 680i, and I instantly went and grabbed my soldering Iron, and did the mods, and confirmed that it worked... then just kinda lost interest, and went back to my main PC.
well, alittle bad luck (killed my P5B deluxe while doing a dry ice test)
now, I am back on my 680i... with alot of free time on my hands, I decieded to give a few things a try.

first, breaking 400+ stable, with a Quad Core.
so, back to the old mod I went.

as of right now, I can say this about this mod.

1. More Voltage is Not Better.
do not, just keep upping the GTL Ref Mod.
you need to find the sweet spot.
which can be really tricky, since it seems the voltage is almost specific to the FSB. for instance 9 x 400 only works at .775 volts... while 8 x 450 is only stable around 1.110 volts.
leaving 1.110 volts, should make 9 x 400 work too right?

nope. doesnt work like that...
you seem to have to tweak the voltage, specific to the FSB and Multiplier your aiming for...

so its a very time consuming mod to say the least.

2. Voltage Range.

for me, anything above 1.2 GLT Ref volts wont even come close to getting into windows. it just will load the blank screen, and never go past that.
so when tweaking, I say aim in the .70 to 1.150 volt range.

3. very picky...

this is a mod that requires time, and patience.
lots of it. lotsssssss of it.
if your the kinda person that just wants to get over with something... dont bother with this mod. so far, its taken me 2 days, and about 5-6 hours of trial and error to get this far, and about 100 OS crashs... corrupting my OS about 5 times, and (sadly) corrupting my back up Hard drive too :(
so be patient, and be methodical when testing this.
theres no quick fix to making this a solid mod.

I'll post more interesting results later.
its 3AM and I am tired as hell now.

And the Article Itself.

http://www.vr-zone.com/index.php?i=4756

Mod Pics.

That's really nice:clap:

What I can't understand is the performance of the quad cores at cinebench. I would expect much higher scores than that. The performance is near to dual core but this is off the topic right now.

Well done.

Still Tweaking.

http://valid.x86-secret.com/show_oc.php?id=201449

456 FSB stable. :D
not bad for a simple 1K VR for $1.

Oh Yeah....

I know the time is slow.
its Vista... its Untweaked for Super Pi... and I have lots of stuff running to monitor my OC's. so its not a speed run, just simple stability test.

yeah vista runs everything 32bit slower than xp! thanks for doing this it makes me want to hold on to my board for a little while longer.

Thats awesome work kunaak, wow 450 on p5n. So without a doubt voltage is consistent with FSB in that for a certain fsb the voltage has to be set at a specific setting? Can you get any 3d at all to run at 450 with quad on p5n??

Quote:

---

Originally Posted by Kunaak

Oh Yeah....

I know the time is slow.
its Vista... its Untweaked for Super Pi... and I have lots of stuff running to monitor my OC's. so its not a speed run, just simple stability test.

---

here is my Vista untweaked time :D :p:


http://img523.imageshack.us/img523/8...hzvistazn1.jpg



[edit] my run was p965 chipset.....it's not a fair comparison as 680i tends to be a fair bit slower

yes, the voltage has to be tweaked with the FSB.

like a few minutes ago, I was testing 458, and I had to get 1.145 volts.
but thats really pushing it, cause 1.150 volts almost always equals a quick freeze of the OS.

I tried to see if I could bypass this freeze, by booting at 1.110 volts, then adjusting the VR to 1.150 volts and higher, but that doesnt seem to help at all.

I can try some 3D but I have 1 problem.

1. I have a block on my southbridge that makes my 8800GTX impossible to use, since the block is too tall, for the card to get into the PCIE slot securely, and it blocks the lowest two, so those arent a option, and I really dont want to rebuild my loop at the moment, as I am more anxious to try the quad core overclocking, then I was to mess with my 8800GTX.

sooo...
the card you see in my pic of the voltage... is a X850XT PE.
not a bad card in itself I know, but not too interesting in 3dmark on vista :(
vistas just slow...

so, if you want them, I will give them a shot, but dont be surprised if the numbers are low.

its vista.... what can I say? haha...

Quote:

---

Originally Posted by Kunaak

its vista.... what can I say? haha...

---

vista is not far from XP in superpi

from my testing ~1.5% and i still haven't tried to disable services in vista yet

Vista and SPi

here's my superpi 1M and 32MB that I did last night.

Quote:

---

Originally Posted by Kunaak

yes, the voltage has to be tweaked with the FSB.

like a few minutes ago, I was testing 458, and I had to get 1.145 volts.
but thats really pushing it, cause 1.150 volts almost always equals a quick freeze of the OS.

I tried to see if I could bypass this freeze, by booting at 1.110 volts, then adjusting the VR to 1.150 volts and higher, but that doesnt seem to help at all.

I can try some 3D but I have 1 problem.

1. I have a block on my southbridge that makes my 8800GTX impossible to use, since the block is too tall, for the card to get into the PCIE slot securely, and it blocks the lowest two, so those arent a option, and I really dont want to rebuild my loop at the moment, as I am more anxious to try the quad core overclocking, then I was to mess with my 8800GTX.

sooo...
the card you see in my pic of the voltage... is a X850XT PE.
not a bad card in itself I know, but not too interesting in 3dmark on vista :(
vistas just slow...

so, if you want them, I will give them a shot, but dont be surprised if the numbers are low.

its vista.... what can I say? haha...

---

Np at all, by all means stick with the x850xt if the least pain in the ass. The numbers don't matter. Just wanted to see if the 3d stability was there on the system side with the mobo. Whenever you get to it is cool. I always start with pi 99%time with new mod/mobo too.
I would bet it is as usually 8m is petty good indicator of some 3d stability.
Highest i could pass 3d at on the striker/p5n was something like 380, and that was pushing it hard so your doing awsome for sure.

@kunaak, the removal of the one resistor in the mod is purely for increase in voltage right? Once the resistor is removed off the board and the mod is removed as well, still works fine?

Quote:

---

Originally Posted by k|ngp|n

Np at all, by all means stick with the x850xt if the least pain in the ass. The numbers don't matter. Just wanted to see if the 3d stability was there on the system side with the mobo. Whenever you get to it is cool. I always start with pi 99%time with new mod/mobo too.
I would bet it is as usually 8m is petty good indicator of some 3d stability.
Highest i could pass 3d at on the striker/p5n was something like 380, and that was pushing it hard so your doing awsome for sure.

---

my typical start up routine with new hardware, or new mods goes like this usually.

Super Pi 1M, 8M then 32M.
Sisoft Sandra CPU math, and bandwidth.
then, any multithreaded benchmarks I know of, Fritz Chessbenchmark, Wprime, Cinebench and Nuclearus.

after that, short runs of 3D.
starting with 2001se.
still my favorite, so I am always curious to see how it effects that first.


anyways...
details if your curious.

Vcore - 1.5
Vdimm - 2.3
1.2 HT - 1.45
NBcore - 1.65
SBcore - 1.6
CPU Vtt - 1.55

Cas 5-4-4-12 (or 10) all others at Auto, to remove the Ram from being a factor in the testing.

Vista32. No updates yet. OS 1.5 days old. No raid.
no sound card installed.
only benchmarks and such installed.
1-2 BSODs so far, lots of freezes from too little or to much GLT voltage though. no Tweaks other then UAC turned off.

X850XTPE using the driver that comes with Vista.
all other drivers downloaded from the net, and are current.

these 450 runs I just did, werent even close to unstable.
it passed them all with ease.
I think its once I get to 8 x 455 where the instability should kick in.
450 is easy. 455 not too hard, but requires some trial and error time.
460, can only boot into windows, havent got it stable for anything yet.


other then this, heres the pics, and two 3DMark Links to check out for youself.

http://service.futuremark.com/compare?3dm05=3075657

http://service.futuremark.com/compare?3dm06=2046678

Quote:

---

Originally Posted by k|ngp|n

@kunaak, the removal of the one resistor in the mod is purely for increase in voltage right? Once the resistor is removed off the board and the mod is removed as well, still works fine?

---

haha...

uhhh haha... well... I didnt think of that.
I dont know if the board will be back to normal once the mod is removed.
not sure what happens to the GTL when theres no resistor there.
I did the mod all at once, and didnt have any intention of removing it.

you know, shamino would really be the one to ask, he came up with the mod, and the credit really goes to him.
I am just the monkey showing some feedback on the mod.

FCG would be another guy worth asking, that guy knows this stuff inside and out. if you really want a tech explanation, hes the cat to ask.

Quote:

---

Originally Posted by Kunaak

haha...

uhhh haha... well... I didnt think of that.
I dont know if the board will be back to normal once the mod is removed.
not sure what happens to the GTL when theres no resistor there.
I did the mod all at once, and didnt have any intention of removing it.

you know, shamino would really be the one to ask, he came up with the mod, and the credit really goes to him.
I am just the monkey showing some feedback on the mod.

FCG would be another guy worth asking, that guy knows this stuff inside and out. if you really want a tech explanation, hes the cat to ask.

---

Damn this sub-forum moves quick now. 2nd page already.
:)..I bet it will be fine with just a little increased voltage left on gtl.
FCG is definitely sharp and knows his stuff. He was telling me all about this mod in detail some time ago before it was more common.
Def. 3d stable though, thanx for that.
455-460fsb x11 3d would be the own on the striker in SLI.

Heres some 450 FSB + 1300 Mhz Ram Action.

it always surprises me how easy 1300 mhz is for these boards.
this is literally, the 7th set of ram I've had that can do this.

http://service.futuremark.com/compare?3dm06=2053387

some pics, and a ORB link for 1300 mhz stable.

hell yah. Nice clkin dude. :clap:
So everything up to 450 is pretty much rock? Just have to have it dialed in right? Ever since using that p5k with the KF I havent been able to look at my striker the same way :D. Really been wanting to do this.
Maybe a suggestion to put this thread in the mod section maybe? it's toast after 1/2 day :)

well. it appears stable.
its certainly not crashing or anything.
I've been running it like this for about 3 days now, with tons of reboots, and testing and all, and only a few times, like 5 outta a hundred boots, has it not booted. so I think I have everything tweaked right for the moment.

only problem I had so far, was just now...
I started Prime, the version with 4 threads.
it was running fine for about a hour, thought I'd take a nap.
3 and a half hours later I wake up, the screen is black, and it wont turn on.
I dont know if the power saving did that, or what...
I been trying to get that disabled, but I think its still enabled somehow.
it does the same thing, when I go to work, and leave my PC on.
sometimes, the power saving feature doesnt seem to work.
I had that problem with vista long before I started this mod.

well... there it is again.
"turn off display after 20 minutes".

I swear, I've disabled that like 100 times so far, it never actually stays off.
thats getting really annoying....

Dual Core Update.

http://valid.x86-secret.com/show_oc?id=202311

this mod even helps Dual Cores.
my board here has always had the ability to hit 530, but never this easily.
with this mod, I just set 2134 FSB in bios, set the ram to sync mode, all the voltages to thier near best places, and rebooted.
worked first time, no problem.
just worked.
in the past, I could occassionally get into windows at 530, do suicide shots...
now, I am at 7 x 533, and its stable.
that almost matchs my P5B deluxe, which I did 7 x 545 stable on that board.
for this board, 7 x 533, is nothing to scoff at.

Quote:

---

Originally Posted by Kunaak

Dual Core Update.

http://valid.x86-secret.com/show_oc?id=202311

this mod even helps Dual Cores.
my board here has always had the ability to hit 530, but never this easily.
with this mod, I just set 2134 FSB in bios, set the ram to sync mode, all the voltages to thier near best places, and rebooted.
worked first time, no problem.
just worked.
in the past, I could occassionally get into windows at 530, do suicide shots...
now, I am at 7 x 533, and its stable.
that almost matchs my P5B deluxe, which I did 7 x 545 stable on that board.
for this board, 7 x 533, is nothing to scoff at.

---

Thats awsome kunaak.
spent about 1/2 day with the mobo yesterday after modding, and so far got 431fsb rock solid for 3d. I know what you mean now by how time consuming this is. Lots of trial and error to find the weet spots for particular fsb's for sure. With all the FSB holes that 680i already has, it makes it that much more time consuming. All the way up from 370 (what the mobo could do stock without gtl mod) I have been trying with the 9x for every bench that passes with 8x, so like you said it's looking good I think for higher multi use.
This run at 431 took about 1.00v gtl. any higher or lower and it wont even post or blue screens. Testing using stock 8800gtx's in SLI:
http://www.overclockingpin.com/431fsb%2006%20kf.jpg

So to start getting closer to 450fsb, how are you handling it? Do you boot in at lower fsb with lower gtl, then raise fsb and gtl acoordingly once in windows? If not, can it be done this way or do you ahve to go straight boot from bios??
Thanx again for the thread and your efforts dude, and to Sham of course for the mod. Lots of work and trial and eroor for sure, but it could payoff huge in the end.:D

Kind of off-topic but @kingpin:
Where did you get your wallpaper?!
EDIT: Nevermind. I found it on DeviantArt. Very nice!
http://www.deviantart.com/deviation/...+age_scale%3A5

Back on-topic:
Nice job guys! I especially like that 100% E6300 overclock. :shocked: What are you cooling your CPU with when you are doing all of this?

Is there a similar, magical mod for the 650i SLI?
Because, my E6300 has trouble hitting around, 480 FSB on my ASUS P5N-E SLI.

I would love a nice 533 FSB. :)

But, other than that, amazing job! I really hate the 680i SLI chipset for some reason, but very nice job!

And, your 7th set of RAM to reach those speeds?! My RAM won't even hit DDR2-950. :(

Good job though! :up:

Quote:

---

Originally Posted by k|ngp|n

So to start getting closer to 450fsb, how are you handling it? Do you boot in at lower fsb with lower gtl, then raise fsb and gtl acoordingly once in windows? If not, can it be done this way or do you ahve to go straight boot from bios??
Thanx again for the thread and your efforts dude, and to Sham of course for the mod. Lots of work and trial and eroor for sure, but it could payoff huge in the end.:D

---

how I do it?

straight from bios.
unlinked.
1800 FSB - 1000 ram usually, though all the way to 1300 work just fine.
set cas, ras to cas, TRP, and Tras manually, and most of the time, set TRC manually too (TRP + Tras should equal TRC so thats how I determine my TRC)
set 8 multiplier.
LDT leave at 5x.

voltages used.

Vcore 1.5 = 1.45 in windows.
Vmem 2.2 normally.
1.2 Ht - 1.45 to 1.55 usually.
NB voltage - 1.65 normally.
SB voltage - 1.6 normally.
CPU Vtt - 1.55.

all stuff posted so far, has been straight bios trial and error.
only yesterday did I start trying Clockgen, and unfortunatly, I still only got 460 FSB tops. which isnt much different then I already did from bios alone.
call me crazy, but I had this hope, that I could boot into windows at 450, and clockgen to 465-475, and benchmark... :p:
always gotta dream big, or why bother right??

well, so far, my absolute max stable, is still 456.
though I can boot up just fine all the way to 470. just never made it into windows.

when going for the real high FSB's, push that NB voltage hard.
for me, getting 460 to post, takes 1.95 volts.
so the chipset voltage is definatly a factor in this.


anyways, I am really glad to see you get to 430+ stable.
thats a huge improvement from 370, a extra 60 FSB is a nice improvement from a little tiny VR and 10 minutes time to mod (and 100 times more time to tune the mod, haha...)

so when you gonna take this to the ORB and do some damage with the extra FSB you just gained?

PS

tell me you werent smiling from ear to ear, that first time it posted and made it into windows at 430... ;)

I was grinning like the cheshire cat in alice in wonderland.

Quote:

---

Originally Posted by Kunaak

PS

tell me you werent smiling from ear to ear, that first time it posted and made it into windows at 430... ;)

I was grinning like the cheshire cat in alice in wonderland.

---

I was grinnin like a felon :D. EAR to EAR. Post 450fsb np now, just need to get it into windows, then work on 3d stability. I also see what you said now about the issues around 400fsb. I think the Fsb holes are making some things a bit trickier to do, but clockgen maybe can get me to where I want to go. Just hope it play nice like this on LN2...we'll see
What bios you on btw?

I use bios 1002.
was thinking of trying beta 1100, but I am not much a fan of changing bioses on this board, since it seems each bios brings up new problems.
1 bios I can't get the Vcore above 1.5 volts. they fix that, then the next bios I cant get the NB above 1.6 volts.
with this bios, I am good on all voltages, most of the time.
when the next official bios comes out, I will give that one a shot.

You could go even higher if you remove/unsolder 8 resistors Vince... ;) :p:
Right now I'm trying to find out the Northbridge GTLREF to mod it too..... :( This Striker by DEFAULT with only 1.86VCore and uninsulated, took my X6800 at ~5340MHz+ benchable in 3D....So I think that it'll do well as soon as I mod it and insulated....we'll see... :)

Vince at 1T and Cas 3-3-3-x, what your Striker can do?.....How high?... Coz this one I got now and unmoded, did only 431MHz.... :(
On ABIT IN9 I could do ~ 445 - 455MHz (chipset depentent) and on the EVGA I could do ~442MHz...

Quote:

---

Originally Posted by hipro5

You could go even higher if you remove/unsolder 8 resistors Vince... ;) :p:
Right now I'm trying to find out the Northbridge GTLREF to mod it too..... :( This Striker by DEFAULT with only 1.86VCore and uninsulated, took my X6800 at ~5340MHz+ benchable in 3D....So I think that it'll do well as soon as I mod it and insulated....we'll see... :)

Vince at 1T and Cas 3-3-3-x, what your Striker can do?.....How high?... Coz this one I got now and unmoded, did only 431MHz.... :(
On ABIT IN9 I could do ~ 445 - 455MHz (chipset depentent) and on the EVGA I could do ~442MHz...

---

Nice! Which resistors and how much higher bro?
what happened to the IN9 benching?? Board is no good?
As for the mem clocks at 1T 3-3-3-x, I think those limits will vary from mobo to mobo. I could get about same on evga/striker maybe slight difference.

Quote:

---

Originally Posted by k|ngp|n

Nice! Which resistors and how much higher bro?
what happened to the IN9 benching?? Board is no good?
As for the mem clocks at 1T 3-3-3-x, I think those limits will vary from mobo to mobo. I could get about same on evga/striker maybe slight difference.

---

With the IN9 I could do ~455MHz max at Cas 3 - 3 - 3 - 3 - 1T - 8 - 17 - 1 - 6 - 2 - 7.8us.....with this one NO GO..... :(

My first IN9 was ROCK STABLE (except the QUADs) till it didn't power up one day.....Code 90 no matter what..... I think that that flucken uGuru chip get's "blocked" somehow and that's it..... :(
My second IN9 did almost the same (power off/power on next day and nothing - no power up)....My third IN9 did almost the same but Digital PWR went down like my Quad GT's...... :(
My EVGA could do ~444MHz at the above (well except 17 - it was 19)......with this Striker I can do 432MHz MAX at the above timmings (21 for 17) and no more.....Now that I have mod it, I'll test it and I'll report back....;)

I decided to go with the Striker for ONE reason.....DIGITAL PWR....IN9 has it and no go with the QUADS......TOO difficult to bench....EVGA has TINY Mos-fets (like the 8800GTXs VGAs).....No go for me PLUS its overclockability sucks big time .....DFI Digital PWR = NO GO ... :D

So I gave Striker another shot and we will see.... :)

About the mods.......You could gain another 20 - 30MHz more (CPU depented)...... :)

UPDATE ..........

DONE...... :D

MAX (without "well" tuning) I'm at 471MHz with Cas 3-3-3-4 and 1T.....I'll upload screenshots a bit later..... ;)


EDIT: Here you go.....This in on my untweaked HD..... :)

http://www.thelab.gr/images/Hipro5/A..._471MHz_1T.png

http://www.thelab.gr/images/Hipro5/A...71MHz_1T_1.png

http://www.thelab.gr/images/Hipro5/A...71MHz_1T_2.png

.....and here's where I want to finally bench with the x12 multi......

Untweaked HD....

http://www.thelab.gr/images/Hipro5/A...51MHz_1T_1.png

....a bit tweaking (LOD, etc.)....

http://www.thelab.gr/images/Hipro5/A...51MHz_1T_2.png


.....and finally on my 3D Benching HD....:)

http://www.thelab.gr/images/Hipro5/A...51MHz_1T_3.png

http://www.thelab.gr/images/Hipro5/A...51MHz_1T_4.png


Bottom line is I think that I'm OK for the time been....Let's move on now.... :D

P.S. If you could follow Kunaak's guide and mod your mobo, you'll be satisfied for sure... ;)

Hipro.

amazing how much one little mod, can completly change these boards huh?

any plans on Quad Core?
seeing Kingpin go from 370 to 430+ stable, was cool.
I am wondering how others would do now too.

Quote:

---

Originally Posted by Kunaak

Hipro.

amazing how much one little mod, can completly change these boards huh?

any plans on Quad Core?
seeing Kingpin go from 370 to 430+ stable, was cool.
I am wondering how others would do now too.

---

A little mod?.... :rofl: I've done 7 of them till now....and believe me, ALL of the GTLREF mods aren't easy to be done..... :D

I've done ALL GTLREF pins/mods 4 of them, Vdroop, VCore, Stabilize and filtered the VNorth and now I'm trying to find out where the heck is that dumn North GTLREF.... :(

Yes that was the reason I've done all those GTLREF mods....coz of the QUAD..... :)

for anyone curious FCG has a great article on the GLT Ref Voltages, and why the voltages arent universal, and do need to be tuned according to the FSB.
his knowledge on the subject is a thousand times more then my own.
I got a monkey see, monkey do, kinda mentality. I see the mod... I do it.
but if you really really want to know why the mod acts as strange as it does... this article will give a great understanding to it.

this quote stood out to me the most, as it was 100% in line, with everything I had experienced with this mod so far.
no one voltage, is universal... just like I said in my first posts..
just nice to finally see I was right, and not imagining it.
when you dont know what your doing, and your learning on the fly as you do it... its kinda scary to stand up and say anything, cause you can easily make yourself into a fool.
glad for once, my experience was right on with how the mod really does go.

Quote:

---

Once you begin adjusting GTLREF voltages (for each core data and address bus) you will find that different bus frequencies respond differently to variations in the voltage. This has to do with harmonics and reflected signals as the FSB changes. The best way to make adjustments is by using a "Clockgen" program from inside the OS to set a previously unstable FSB frequency while simultaneous moving the GTLREF voltage points. Memory or data intensive applications (such as Super PI) will stress the system the most. This inherent variation in system response and stability while passing through different ranges of FSB is most likely the reason that some users experience what they describes as "FSB holes." Just because the FSB signal timings are tuned for 350MHz and 450Mhz doesn't necessarily mean that they are adequate for 400Mhz FSB. Figure 5 illustrates this concept graphically.

---

http://www.thetechrepository.com/showthread.php?t=87

anyways, give the article a read, if your interested. its quite interesting to finally see whats really happening when you use this mod.

but after reading it... it only makes me even more curious to tweak the board even farther now, since it kinda clues me in on somethings.

Did the mod exactly as it has to be done and the result now is that system won't initialize the CPU. All the voltages are OK - vCore reading 1,26V, vRef = 0,87 (originally 0,77; tried up to 0,96 and down to 0,66) - still no go. Unsoldered the variable resistor and put new one in the place of the originally installed on the board (214ohms) I had to unsolder to do the mod - the Striker still refuses to initialize the CPU. Any ideas what could have happened, or all I can do is get a new board?...

One more question - with the removed 200ohms resistor and soldered 500ohm, I couldn't up the GTL Ref voltage more than 0,96V, however I read in this thread people are using 1,1V... how is it possible?

EDIT: RMA'd the board, next week getting a new one, hopefully. Getting 3220 pretty soon too, so I will definitely need the mod for 400+ FSB and I need some opinions what would have gone wrong last time; is it mandatory to remove the resistor, as Shamino gives instructions, isn't there a possibility for the mod to work without doing so?... And is 500ohm VR fine - as I said, I couldn't get more than 0,96 vRef, (I saw Kunaak uses 1K, maybe this is the way to go).

hi,

for those who have successfully completed this mod, what thickness tip did u use on the soldering iron? is it worth getting a 0.12mm tip for my 25watt iron, or is this overkill? and where did u find was the easiest place to earth it?

The RMAed board returned to me, moded it and it works fine now, only when I measure the vCore with DMM I get 1,6V when 1,9V are set in BIOS. Strange, didin't have this issue with my previous Striker... Next week, probably, will try with a quad and see how the mod helps, until then I will play with a crappy 6700, that won't do even 3,75GHz stable...

About the grounding - I prefer to use screw hole, there is one where the heatpipe is curverd, just a few cm above the resistors, participating in the mod.

=( I'm stuck at 350FSB...

When I FIRST got my G0 Q6600, I was able to do 3.33GHZ, without even trying. Then I did some undervolting, just to see how low of voltage I could go on stock speeds. Now I'm trying to overclock, it won't let me get past 350FSB.????

And if I'm reading this right, I have to take a soldering iron to my motherboard?! =| =|

My settings
3.15GHZ Q6600 G0, 1.3875v (probably over volted)
P5N32-E SLI 1103 BIOS, NB 1.5, SB 1.55, CPU VTT 1.55, 1.2VHT 1.4
OCZ DDR27200 2GB, 888MHZ, 4-4-3-15 2.2v
edit: my idle temps are 30C/26C/27C/27C, load temps 43C/43C/40C/40C

Any suggestions besides soldering the MB?

Quote:

---

Originally Posted by treatmentx

=( I'm stuck at 350FSB...

When I FIRST got my G0 Q6600, I was able to do 3.33GHZ, without even trying. Then I did some undervolting, just to see how low of voltage I could go on stock speeds. Now I'm trying to overclock, it won't let me get past 350FSB.????

And if I'm reading this right, I have to take a soldering iron to my motherboard?! =| =|

My settings
3.15GHZ Q6600 G0, 1.3875v (probably over volted)
P5N32-E SLI 1103 BIOS, NB 1.5, SB 1.55, CPU VTT 1.55, 1.2VHT 1.4
OCZ DDR27200 2GB, 888MHZ, 4-4-3-15 2.2v
edit: my idle temps are 30C/26C/27C/27C, load temps 43C/43C/40C/40C

Any suggestions besides soldering the MB?

---

Try booting at 1600 - 1680 within bios, Linked and at 1:1....

Do these mods apply to the P5N32-E SLI Plus??

I ask becuase I am certainly experiencing similar FSB issues with G0 6600 Quad.

Currently using latest ASUS 0805 BIOS.