45nm can't use speedstep technology on P5W. Maybe only in newer bioses.
To change CPU voltage try to clear CMOS by jumper, maybe after it, you will see Vcore options other then Auto.
Printable View
45nm can't use speedstep technology on P5W. Maybe only in newer bioses.
To change CPU voltage try to clear CMOS by jumper, maybe after it, you will see Vcore options other then Auto.
yolanda, could you be more specific? how do i clear CMOS by jumper??
everytime i flash the bios, there's an option to clear CMOS, but that didn't work :(
thanks!
Hi,Quote:
Originally Posted by jpmermoz
you should try clearing CMOS moving the jumper on the motherboard. The P5W Manual will tell you how to do ;)
Hi!
You can try this tool RMClock utility here : http://cpu.rightmark.org/
My Q9450 under vista has a VCore V1.15 @ 2.66GHz with this tool. Speedstep works too.
ok, i cleared the CMOS just as the p5w manual says, but i still can't change the vcore option :SQuote:
Originally Posted by PaVi90
i downloaded the rm clock utility but i can't change vcore :SQuote:
Originally Posted by teaday
Ok, that's too bad :-(
My P5W is a rev 1.03G
I have an Asus P5W Deluxe revision 1.4g with bios 2801.
CPU is an Intel E8600, running at stock speed with Zalman CNPS7700CU.
CPU fan at max speed.
My idle temps are 38° for both cores.
Messured with speedfan en Realtemp (latest versions)
Are these "normal" temps?
I have those temps in my E7200 with stock cooler and stock speed.Quote:
Originally Posted by Ragdd
I think it's normal
Last day when I booted my computer, it didn't show any picture. Reset the BIOS helped, but now I can't change the vcore to get my old stable clock at 3.7GHz back. I tried, but whenever I change the vcore to anything else than auto there's no picture when I boot up. Any ideas or is something wrong with my mb? Is it possible the CPU have taken damage from long time with higher clock and vcore? The temps were alright, but maybe the vcore did some other damage?
All, I now have a 8500 C0 stepping with 2801 BIOS. I'm experiencing the limited range of voltage options for the CPU (1.45 -1.70V). I have a 1.02G (1.65vmch version). I've also noticed that with 2801 in the section for selecting CPU options, multipliers etc, under 2801 it gets the multiplier info wrong, from memory displays a multiplier of 73 ?!? Anyone else notice this ?
Hopefully a later non-beta BIOS will fix these issues.
One thing's for sure: if more vCore isn't the resolution to the instability issue I have with the Q9450, I'll rethink my position on keeping this motherboard until I move on to a Nehalem CPU in a few years. It'll be hard to get a new MB with two IDE ports for my 4 devices, but maybe not impossible...
Does anybody use an aftermarket GPU cooler with this board? How have you plugged the cooler in?
I have an Akasa Vortexx Neo on my new 4850 which has a three-pin connector on it. Therefore I have to plug it in either into the PSU directly or to a P5W fan header.
I have it plugged into the 'CHA_1' connector at the moment, I've also tried PWR_2. Basically, 3d intensive stuff seems to crash after a few minutes when graphically intensive games are run. Speedfan doesn't seem to think the fan on the cooler is spinning up much, and judging by the sound this seems to be correct.
How can I get the cooler to spin up properly, as when I run GPU benchmarks, the temp just seems to go up and up :(
I'd say you got a faulty aftermarket cooler, I've had all fan connectors on the mobo populated (mostly by 120mm fans), and all fans worked properly.. It's possible you mobo is faulty but I'd try a different fan first ;)
Just tried it with another one of exactly the same kind (we bought two 4850s with these coolers attached off a guy who has upgraded to a 4870x2) and it's the same thing.
It just doesn't seem to have the grunt to spin the fans up to a decent enough speed to stop the cards getting hotter! This is plugged straight into the PSU - should this not mean the fans are on 100% all the time?
Apparently these are highly rated coolers for these cards as well :(
Why would they not seem to want to spin up when the card gets hotter? Is it because it's not plugged directly into the header on the GPU anymore or something? Or are these sorts of coolers just not supposed to work like that?
Man, I just upgraded to play Crysis better - and it runs beautifully until I get display corruption and everything crashes!
it just blue screened. Damn it.
Dumb question: are you using fan power ports that are controlled by the BIOS, where disabling the fan control would let them spin as fast as possible?...
cheers for the reply - at the moment the fan is plugged via the three pin to molex converter supplied with the fan into the PSU.
I did have it plugged into the P5W directly though, it seemed to have the same problem. I had the 'Q-Fan' controls for the fan header and 'AI Quiet' turned off in the BIOS :(
Graaah.
radlord, there are options for fan control in bios, don't remember what or where (don't have my p5w any more), but try turning them off and check in SpeedFan that they are at 100%. And of course, mobo wouldn't know that your videocard is overheating, it won't bump up the RPM on the coolers...
Hello,
has anyone a Q9450 on P5W? How does it work? :)
Underclocked, it works fine. Run it at normal 1333 speed and I believe you will have spontaneous reboots. Best guess: not enough voltage provided by Auto setting, and 1.45V which you could pick from manual settings is a touch too high.Quote:
Originally Posted by PaVi90
At least, nobody seems to WANT to set it that high unless they have free replacement CPUs available...
uhm... wich BIOS do you have currently? Is a C0 or C1 CPU?Quote:
Originally Posted by mikepaul
You tried 1.45V, your quad stable or not? It's not so high, because real voltage will be about 1.41-1.42V.
About latest 2801 BIOS: my system becomes Unstable with it! It freezes about 2-3 times per night at FSB 425 and about 1-2 times in day at FSB 400, then I play 3D games. All tests (Prime, Linpack) are fine. But in game sometimes system crashes. I added voltage to CPU, loaded setup defaults, used more powerfull 600W power supply, but nothing helped. Then flash older 2704 bios and now everything is OK. Maybe it happens only with my config, I have Geforce 8800GTX, used 3 latest driver versions.
Going on memory (will correct if needed): BIOS 2801. C1 stepping...Quote:
Originally Posted by PaVi90
If I had another $300+ to throw at a replacement Q9450, I'd try 1.45V.Quote:
Originally Posted by Yolanda
So far, I haven't had it...
Mm a lot of systems from brand manufactures uses QX9650 CPUs at 1.5-1.6V with factory overclock at ~4GHz.
So you only need normal cooling to cool additional heat power. However you have a warranty on CPU...
I Googled and found some posts where 1.45V is the max for a Q9450, while others said 1.35V or so was as far as it should go.Quote:
Originally Posted by Yolanda
I tend to run my PC maybe 14 hours a day on work days, so I'm not SURE that the Freezone Elite will compensate enough for the overvoltage to keep the CPU going for the next few years.
And I don't think the warranty covers setting the voltage too high...
can we expect a BIOS update to get Speedstep working on 45nm then, or is it technically not possible?
I've just installed E7200 and so far it works great. I'd like to ask you what's the maximum safe voltage. About that Q9450 stability issues, you can try 1,45V just for a few minutes of testing. I'm sure that your CPU won't die. Intel doesn't make any troubles when it comes to RMA.
There are two aspects of changing settings on any board; hardware and software. The hardware must first be capable of selecting a finite range of values and then the Bios must be written to allow each of those values to be chosen. I have a rev. 1.03g board, (and an e6700) have tried several bios(s) and have always had the minimum vcore range of 1.225 V. I think someone on this board reported Rev. 1.02g also allowed a < 1.45V selection. The question is - what do versions prior to 1.02g allow. If it's truly a hardware issue, then a new Bios is not the solution.
You have 1.225V with 65nm CPU, but with 45nm CPU only Auto (nominal) and 1.45 and higher can be selected.
Always was a unwritten rule in overclocking, what voltage can be rised upto 10-15% above nominal without troubles :) So nominal is 1.225-1.25V and you can set to ~1.4V.Quote:
Originally Posted by tomcug
Yolanda, thanks for your answer. At the moment I'm running my CPU at 3,8GHz and 1,36V. I'll try to reach 4GHz tomorrow. P5W DH Deluxe is really an awesome motherboard and I'll keep it until Nehalem :).Quote:
Originally Posted by Yolanda
Nice. I take it you have a 1.04 revision of the P5WDH?Quote:
Originally Posted by tomcug
That's right, my motherboard's revision is 1.04G.Quote:
Originally Posted by Tamerlane
Did you get ur e7200 to run at 4ghz yet? What FSB is needed to get there?Quote:
Originally Posted by tomcug
Yes, I managed to overclock my E7200 to 4GHz. I had to increase FSB speed to 421MHz to reach that frequency. Right now I'm not sure if it's fully stable, but I'll check it soon.Quote:
Originally Posted by anzial
421FSB required to get 4GHz on E7200 and E8500. But only from 423FSB BIOS will print at startup about your nice "4.01GHz" :)
Could never get much over 400mhz on my board... dunno why
I had to increase FSB Termination Voltage to 1,3V and NB Voltage to 1,65V to reach frequencies over 400MHz.Quote:
Originally Posted by anzial
However 45nms need large voltage boost to get to 4GHz and over.
My E8500 works at 4GHz at 1.4V, at 4.2GHz at 1.5V, but 3.85GHz requires only 1.32V.
Try to test by Linpack, Prime heat 4GHz CPU upto 70-75 by RealTemp, but Linpack get upto 85-88...And it's really hot, ASUS Pcprobe reports about 60 degrees at this state.
That's right. That's why I'm usually running my CPU at 3,8GHz. I'm running it at 4GHz only for benching.Quote:
Originally Posted by Yolanda
I'm have watercooling, so CPU temps were never my problem. Mobo just didn't like going over 400mhz
You have to increase FSB Termination Voltage and NB Voltage. It helped in my case.Quote:
Originally Posted by anzial
I did that to get to 400mhz... anyways, I don't have that mobo any more, so it doesn't matter much now :)Quote:
Originally Posted by tomcug
I just picked up a Q6600. If my bios is already up-to-date what steps do I have to take to upgrade from a e6600 to a q6600 on Vista Home Premium? Do I have to reinstall or can I just replace the processor?
You might have to reactivate vista but that's about it. CPU will be automatically recognized by windows, there are no special drivers fo Q6600
does anyone with a 45nm proc have EIST still available in the BIOS?
Thanks guys. I have my new Q6600 OC'd at 3.10 with 1.325 volts I haven't done any stress testing yet but was wondering what kind of differences would I see between 800mhz memory at 5-5-5-15 ratio 5:4 and 667 mhz memory at 4-4-4-13 ratio 1:1?
Also, the temps look to be pretty good in the low 40s.
Guys, I've got some troubles and I really need your help. Despite the fact that my CPU is rock stable at 3,8GHz, memory controller is unstable at that 400MHz. Maybe I should install 40mm fan at stock cooler. I don't think it's caused by too small MCH voltage, because raising it doesn't help. I'm pretty sure that my motherboard is simply overheating.
About 8X00 C2D CPU and P5W DH Deluxe motherboards with EIST issue:
Found on Asus Forum
http://support.asus.com:80/faq/faq.a...Language=en-us
Problem
I notice that this motherboard starts to support Intel E8000 series CPUs since BIOS version 2704, so I then upgraded BIOS to this version. My E8300 CPU runs well, however, neither EIST nor C1E function is shown in the BIOS, which according to Intel's website is supported by this CPU. How do I fix this problem?
Answer
P5W DH Deluxe is based on Intel X975 chipset, which limits FSB at 800MHz. With ASUS unique overclock technology, this motherboard is able to be running at FSB 1066MHz. When Intel launched FSB 1333MHz CPUs, we once again overclocked FSB to 1333MHz to support them. However, considering system stability, we do not support FSB and multiplier modification upon 1333MHz FSB platform. That's why EIST and C1E functions are disabled on your PC.
Tried out the E5200 on my 1.02G 1.85v MCH mobo and lowest manual Vcore is also 1.45v, looks like proper 45nm support badly needs a new improved BIOS from Asus. :(Quote:
we do not support FSB and multiplier modification upon 1333MHz FSB platform. That's why EIST and C1E functions are disabled on your PC.
I've just fixed my issue with memory controller :). I've changed timings to 4-4-4-12-5 and I've raised MCH voltage to 1,65V and memory voltage to 2,2V. Now my system is rock stable :).
How much memory do you have, and what brand is it?Quote:
Originally Posted by tomcug
I've done some more accurate testing and it seems that I only had to increase memory voltage to 2.05V. My memory is 2x1GB Team Group Xtreem Dark 800MHz CL4.
Does anyone know how to get the attention of whoever writes the FAQ? If we all Asked the same Question Frequently enough, they might bother putting out an answer as to why they mistreat the owners of older versions of the DH Deluxe...Quote:
Originally Posted by MacClipper
I seriously doubt you gonna get any decent support for P5W DH, it's too old from ASUS's point of view, too few users that try this 3-year old tech with the latest and the greatest CPUs.
The current state of E5200 support is most unfortunate cos the 12.5x multiplier equipped E5200 is a great cheap CPU to pair with the similarly FSB impaired P5W, lots of multipliers to play with.
Been a awhile but I have a new question. I searched but couldn't find an answer. I found a Q9550 for $328 and wanted to know if it'd be worth it to use with the p5 w dh or should I just build a new system. Is it a hassle to fine tune the quad cpus with this board?
The board is going strong but I have noticed a lot more slow downs when multitaksing with Office 07, Photoshop, Browser and a Music running at the same time. If I have a folder with my photos and it set to Thumbnails it really messes with video performane or audio sound. I figured a new processor and change the RAM timings may fix it.
I'd have to say you should expect hassles with the P5W and a 45nm CPU if your MB is older, but you should try it if you don't really want a new MB. If it fails to work right like my Q9450, you'll have had the time to scope out the new equipment...Quote:
Originally Posted by tktnuri
Same boat as you running Rev1.03g e6600@3.6ghz for over a year now. wouldnt mind a Q9550 i called Asus in UK and asked if Q9450 would work on rev 1.03g they called me back after a few hours and said YES.Quote:
Originally Posted by MacClipper
Q9550 Latest bios 2801
Any idea if I can use Q8200 in this mobo?
you can try, and it'll prolly work, just don't expect any overclocks with this setup. 975 is getting old...
I was wondering if anyone can assist me with my motherboard problem. I recently tried updating my bios with the asus update tool for windows to the latest build of the asus bios for my
P5w-DH Deluxe motherboard. But the problem is when my pc boots right now it says Block compatible version .035 searching for bios, I thought this has crashfree bios 3, so I inputted my
support cd b ut it tells me it can't find cdrom drive so i insert the latest build of the bios and put it in a usb flash drive then the computer reads the P5WDH.rom file but then
it says searching usb device again, it just keep searching and searching then it searches cdrom drive then floppy. anyone got any advice what to do or how to resolve this thanks
Write p5wdh.rom bios file on floppy and try.
Yolanda, why are you running your memories at 5-6-6-18 timings? Did you have to enable SPD timings in BIOS for these clocks to be stable?
Yes, by SPD. If I set anything not by SPD, system can't even boot to BIOS :)
Yolanda, I have the same issue. I need SPD to boot @ 400 fsb.
I'm planning buying an E8600 and I would love to run it @ 4 ghz but right now I'm unable to achieve rock solid stability with my old E6600 @ 400 fsb (using a lower multiplier)...
Could you please share with us what settings and voltages do you use?
(FSB Termination, ICH, etc. ?)
Thanks!
:poke: memset ;)
These are my voltages:Quote:
Originally Posted by firestarter.pt
FSB Termination Voltage: 1.3V
MCH Chipset Voltage: 1.65V
ICH Chipset Voltage: 1.05V
I'm also using custom timings, but trying to boot at speeds over 400MHz is always a failure. In that case I have to use SPD timings.
At 400FSB I use all voltages at Auto except CPU, HyperPath3 Disabled, Timings by SPD, PCI to 33.33, PCI-EX to 100 (but they also can be set to Auto), Perfomance Mode to Turbo. CPU Thermal Option and Memory Throttling to Disabled (but they can be Enabled, can't see difference), PEG Link, Latency and Slot Power to Normal.
At 421-423FSB I set MCH to 1.65V and also increase CPU voltage from 1.33 to 1.425V.
Yolanda, did you notice any performance improvements with Performance Mode set to Turbo?
No, I think it maybe now change some timings or other settings only in Auto Overlocking. However then I have 65 nm CPU and old 22XX-24XX Bios, it speeds up SuperPI result for severel percents, now I can't see any diiffrence, but what is way I use Turbo from that times, and without Turbo I can't also get higher overclocking.
I'm almost SURE That Turbo changes some settings if everything is set to auto.
On the old BX chipset, Turbo and F1 and Speedracer, besides standard, were the settings available, and each one affected certain subtimings that you couldn't change manually. But those worked with manual cas latency settings; setting F1 stopped my board from booting at any higher FSB. Pretty sure Turbo does something similar here; maybe related to auto settings, probably unimportant. I never saw any difference with a 65nm cpu, at least with 1M spi....
Thank you all for your input. Now I can boot with no problems @ 400 fsb and it seems stable but it's still early to say if it's 101% or not :)
Do you think this MB can be rock stable @ 400 fsb with a E8600 for 24/7 usage? Or shall I wait a bit more and then upgrade both MB and CPU at the same time?
Can't say why it wouldn't be stable with E8600. I'm running my board at 400MHz with E7200 and I've never experienced any problems. My board was stable even at 450MHz, but I can't say the same about CPU :) It's still fast and great board. I want to buy second HD 4850 and CrossFireX both cards. What do you think about it? I usually play at 1600x1200.
Yolanda I have a question, I did the math on your FSB and CPU speeds and it sure looks like you are using a 9.5 multiplier on your e8500. I too have an E8500 in a P5W DH, and under CPU ratio I can't go above 9. I am using BIOS 2801 and thus far have only OCed to 367 FSB = 3.3 gHz CPU @ 9x. Did I miss something in the BIOS or do I need 2704 like you have?
PS: I have 1.04g and I can set vCore to < 1.45 (I think as low as 1.8v but not sure at the moment, low enough though)
Just disable CPU ratio modification in BIOS and you'll get the 9.5 multiplier.
Sure, disable CPU ratio modification or simply load setup defaults, and E8500 will work on it's nominal 333FSB * 9.5 = 3.16GHz.
But how you can change CPU ratio on non-extreme 45nm CPU? P5W doesn't support EIST on them and so CPU ratio modification can't work. If I choose anything from 6 to 9, system even can't boot to BIOS, and I should reset board by CMOS clear jumper.
hmmm i have ratio modification disabled and it still shows up as 9x. also, the POST message is misreporting my cpu freq as "432mHz" (but CPU-z reports it correctly, or at least what I was expecting) so I'm going to reset my CMOS and try again . . .
I thought you couldn't unlock the multis on non-X cpus and in any event Asus won't support different multis on this board at the FSB speeds needed to run the 45nm CPUs ??
I searched and could not find an answer to this (on other sites too), so I apologize in advance if someone has answered this.
I recently replaced an e6600 with a q6600. My old e6600 was able to do 400x8 with a little extra juice (1.4v). The q6600 shows potential, currently running rock solid at 333x9 at stock (1.28v) with my 4gb dual channel sticks at tight 4,4,4,12 timings at 1:1. Unfortunately, I cannot even get close to 400 fsb because I cannot get the multiplier to change, meaning my DDR2-800 RAM is running at DDR2-667 instead.
I can set it down to 8 and save the changes to the bios, but as soon as it reboots, it resets back to 9. I strongly suspect it as an issue with the bios because I AM able to drop the multiplier to 8 if I leave the fsb at 266 (i.e. I can boot at 266x8 and have it stick), proving my CPU can accept a lower multiplier. As soon as I raise the fsb past 266, I get the multiplier issue where it self-resets to 9.
After doing a lot of research for a week and seeing some sites where members claim that any bios versions past 1707 have issues with dropping the multiplier, I tried flashing to 1707. Unfortunately, since I have a more recent rev. g0 of the q6600, 1707 does not recognize the CPU and gives me a microcode violation. I can go past that, but that issue is annoying itself.
So I tried bios versions past that included 2206, 2704 and 2801, with no luck.
I have all the fix settings suggested throughout this thread otherwise (i.e. look on the very first post, and it is a mirror image of what my bios are set to). So I am not sure what else could be the issue other than bios itself.
Am I out of luck?
I don't know why you want to go down with multiplier. If you don't want to overclock your CPU to higher frequencies, then simply leave it at 9x. You can set your memories to run at 834MHz.
I am trying to overclock to clarify.
If I go down to 8x, I might be able to run at 400x8 at a relative low vCore, at least that is the goal. Then I have greater fsb bandwidth, a higher overall clock of 3.2 GHz, and can run my memory at 1:1 at 800MHz. Right now I am stuck at 333x9 (3.0 GHz) with memory at 667MHz (I know my 9x multiplier wall is somewhere around 350 fsb as I could boot into windows at 360 but was not prime stable).
I highly doubt my memory can go too far past its 800MHz rating as it is a set of Corsair 2x2GB CL4 DHX, which are known to be a bit sensitive.
8x multi is likely to hit a worse fsb than default multi..
Our board overclocks with dual cores upto ~400+ and something more by FSB, but with quad cores overclock is much lower. So 333*9=3GHz is rather nice. Tried to set memory to 834MHz with your FSB333? Сan it work? Or can work with lower timings? because usually higher memory frequirency is better than timings.
After it, you can overclock from 333FSB to ~340-360, and get extra MHz, and also get extra MHz to your memory. If memory can run only at 667MHz with 333FSB higher overclocking will also overclock memory to 680-720MHz, not so bad.
Lawrywild, why do you say that? Just curious? I mean from what I have done with my old dual core e6600, 400x8 was way better than the 350(ish)x9 that I could get both in terms of final frequency and in terms of the few benchmarks I ran at that time. I know I am working with a quad now, but I just find my problem odd. My issue is not with the chip itself, but why on earth does the motherboard reset my multiplier. It gives me the option to lower it, but then it won't stick.
Yolanda, thanks for the suggestions. That is my plan if I cannot figure out what is with the multiplier. As just stated, I am confused why it allows me to lower the multiplier but won't stick unless the fsb stays at stock. To me that is a bios flaw, not a board or chip flaw, which is frustrating, since that means Asus dropped the ball somewhere and is probably not going to fix it.
Anyone here have experience with the p35 based Asus P5K-EPU (or alternatively the P5K-Deluxe)? Do those have the same multiplier issue? I might be able to get an EPU for cheap or even free, or a Deluxe for just a few bucks. But I am reluctant to replace an otherwise great board over a bios issue.
OK, I was reading about Real Temp 2.70 today and decided to try it. It reads 10C lower than Core Temp with no tweaks, and I never tweaked Core Temp either.
I like it saying 39C instead of 49C, but is that valid?...
Please help, System won't post because turn the OC too high after upgrade BIOS (2801). Please help me to resolve this without to use CLRTC...because I don't want to remove my Video card GTX 8800. Is there any ontherway to bring back the system?
In the Menu book's ASUS, it said I can use C.P.R ( CPU Parameter Recall) But I don't know how to use that function. Please Please help me. Thanks very much. I'm killing myself.:shakes:
Ntvh, load BIOS default setting by using CMOS jumper. You'll find instructions in your motherboard's manual.
Thank, IS there otherway by not remove that jumper CMOS because to do that I have to remove the Graphic card on the its way. Is there easy way than to remove CMOS? Thank again. and this is for what I read from the user guide on page 2-26
" You do not need to clear the RTC when the system hangs due to OC. For system fail due to OC, use the C.P.R. (CPU Parameter Recall) feature. Shut down and reboot the system so the BIOS can AOtu reset parameter setting to default values "
How, where???
Thanks
If you overclock too much and your system doesn't boot up, then you have to simply cut the power. After few minutes you turn on your computer and it'll run at stock settings.
If it can't help, you should reset BIOS by CMOS clear jumper. Why you can't do it? You don't need to remove your 8800GTX from the case to make it.
i tried not to remove but it on the way. I could not touch it..so have to remove the card. Anyway, it is done, now it is running 3.5 @ 390 with the DDR at 983. Will try to push it to 400. Thank you for all of your help:D
BTW: never can set the Multi to 9.5. Any suggestion. THX
yeah i cleared my CMOS and set CPU ratio modify to disabled and don't have the 9.5 multi
i am running at 367 FSB with 9x and 734mHz DDR and things seem to be running ok otherwise
my BIOS is 2801, i am wondering if I should try 2704 since I think that is what Yolanda is using to get 9.5x, unless there is some other trick I'm missing
MaximumPig, I've got E7200 which has 9.5x multiplier too and I don't have any problems with it and BIOS 2801. Post a CPU-Z screenshot.
here is my picture of CPU- Z. Thanks
Ntvh, you won't set 9.5x multiplier with E8400. Your CPU has its multiplier locked at 9x.
Thank you SIR.
Yolanda, What can I do to OC the CPU to 4.GHZ? Please help.
Here is my set
Bios 2801 ( should I go back to 2701?
MAX CPU Freq: 378
DDram: 978
Perform Mode: Auto
PCI Express: Auto
PCI Clock Syn Auto
Mem Voltage: 2.00v
CPU Volt : AUto
FSB Term : Auto
MCH Chips Set: 1.64v
ICH : Auto
Digital...: Disable
C1E: Disable
EIST: Disable
If you need any info please ask. Thanks
You need FSB 445MHz to get 4GHz with E8400. Try to set MCH voltage first to 1.75V, then if it can't help, to 1.85V. CPU voltage to ~1.4-1.45V. Later add FSB voltage if nothing is help. RAM speed to lowest mode, because it can limit overclock at first tryes. However 1.85V on MCH is rather dangerous for 24/7 without extra cooling.
First, I advice you not to try such high overclocking, but to test only FSB 400MHz with memory at 800MHz (FSB:RAM = 1:1). Your CPU will work at 3.6GHz, it's rather fast. This frequirency needs only to add CPU voltage a bit.
After test something with FSB ~410-420. There setups are ready for 24/7.