• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

My EP45-DS3L + E8400@4Ghz + air cooling settings

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
When you say it wasn't stable, do you mean it wouldn't boot or you mean it won't pass stability tests? I think I was able to get to around 4Ghz on this board just by changing vCore to 1.4V and the FSB to 445 and leaving everything else on Auto. I don't recommend keeping it that way indefinitely though but maybe it'll get you a start. Also try moving vCore and FSB up slowly in small increments, that works for some.
 
just for reference the 8400 should use much lower volts to get to 4.0 as opposed to that q9550. quads stress the mobo much more and will probably need some more volts to get high.

take it slow and find the max fsb you can get @ stock volts and bump the cpu voltage up then find that max fsb again and repeat until your comfortable or the temps arent safe. just be sure to keep an eye on those temps when stressing are around 60°C .

I've never had any luck just mashing the fsb real high and praying for the best. just take your time.
 
By stable i meant it would give me blue screen errors or freeze unrecoverably after everything I tried.

Thanks for the advice guys I will try this more gentle approach of upping the fsb and vcore only before tinkering with the finer settings. Any ballpark voltage suggestions for my quad?

Ps Smitdogg nice clock, I'm envious
 
Last edited:
Edit: my BIOS resets itself if I set MCH Core to anything past 1.3. It will run fine the first time I set it but after a couple of resets, it defaults my whole BIOS. It looks like I might have to run my vCore up to 1.3875.
Instead of raising the Vcore to try and compensate, I think I might try and find a solution to the BIOS reset issue w/ vMCH > 1.3v. It could be a BIOS issue, whereas a different revision may not exhibit the same behavior.
 
Hello everybody! i am new in this forum, and i wanna thank you all for this information! I will write a longer post as i am having big issues with similair system as smitdogg. Here we go:

My system spec:

e8400 q813A168
EP45 DS3L - i am using the new f9D bios from gigabyte site
also tried all previous bioses
4gb mushkin ddr2 1066
9800gtx
600w psu
----

Now what i am trying to do is have 4ghz for 24/7 playin games, encoding videos and stuff like that. I am having huge problems. I was tryin the settings above from smithy, and system would not even post with 1.37V ... I only get a beep and black screen, then system turns itself of, and back on and the bios restarts to defaults. Now i noticed some weird things:

I run @ 445 x 9 = 4.0 ghz @ 1.550 V in bios (1.500 idle , 1.47V on stress because of board Vdrop) i cant raise mch voltage even by bit because it wont post, (ram @ 1068)

I run @ 500 x 8 = 4.0ghz @1.550V i can raise MCH to 1.3V without problems (ram = 1000mhz caus i put it down to 2.00B in bios latch settings)

1. My batch number claims to be good on inet forums, people say they need 1.37 V for 4ghz on my batch number and cpu. I need 1.5V to get it stable on prime95, anything below 1.5V results in INSTANT resatrt when running 2 min of prime95, however i can run most games without any problem.

2. My motherboard reads default cpu VID as 1.225V , but on my previous motherboard p35c ds3r it was 1.1125 V which is also what shows core temp and other programs as my cpu default VID. So my motherboard ep45 ds3l read s wrong VID. Does it read the same on your bios smitdogg? Please can u check and tell me would help alot.

3. It bit impossible that i need 1.5V on this motherboard , and i was running p35c and i only needed 1.37V (which is also what its written on forums around for my batch number and its normal) and it was perfectly stable.

Now i am stuck , my system is stable on 4ghz, but 1.5 V?!?!?! its like 72 C on prim95 after 2 hours.. i think its bit too much,.. and if i think that i changed my p35c for ep45 because i tought its gonna be better?? I can only say gigabyte fails with p45.. as it sux, i sohuld have bought abit ip35 pro.. wouldnt have these problems now .

I tried almost every setting, puttin all rest voltages on auto and raising only vcore, didnt help at all.

One thing i noticed up on your screenshot smithy is that in Everest cpuid it shows u information about motherboard chipset and stuff,.. when i open everest it doesnt show any information and it says "unknown motherboard" under brand and chipset in cpuid those spaces are blank?! I am using the latest everest 4.50 ?? Do you have some newer version? Or i didnt install some drivers or something i am missing?

Help me guys, please i am rlly gettin mad, and about to return board for my old board again ;\
 
Last edited:
Instead of raising the Vcore to try and compensate, I think I might try and find a solution to the BIOS reset issue w/ vMCH > 1.3v. It could be a BIOS issue, whereas a different revision may not exhibit the same behavior.

I agree with you 100% that it could be a BIOS issue, in fact I was thinking that yesterday when I was testing. It'll be a matter of waiting though before we can find out because I'm using version F7 and every prior version is so buggy that they make my PC BSOD constantly when I have ForceWare installed. So it's wait for F8 or nothing.

Edit - F8 just came out ;-)

It looks like my Vcore requirements have changed in this BIOS update, it won't even boot from my last settings, which is strange. Lots of testing to do...
 
Last edited:
smitydogg there is already f8 bios on ga page.. can u answer me the question i asked u please? would be really thankfull
 
OK I don't know if I have all the answers you want. First off are you talking about raising Vcore to 1.5V+? I wouldn't run mine that high even temporarily. I don't know if you got a bad chip or what.

I'll have to check my BIOS later to find the VID but I'm almost 100% positive that it's ~1.225.

Re: Everest - try installing SpeedFan and let it run for a while. Weird solution I know, but it fixed some behind the scenes stuff for me.

Make sure your TJ Max is set to 95C in your testing programs for the E8400.

What we were saying before is go up slowly. Like try turning your Mhz to ~3.4Ghz, barely raise vCore past default, run it like that for 8 hours, raise them to 3.6G, run that for a few hours, etc. Going up slowly works for some people, and dude anything past 1.4 Vcore on air cooling is... Well I wouldn't do it for very long.
 
Hello man! Thx for answers. Well about the bios.. i was checking gigabyte homepage for bios regulary every day.. 3 or 4 days ago there was suddenly f9d bios on gigabyte page, and it said 1333+ memory support, now 2 days ago i checked again and there wasnt any f9d anymore only new f8 with same description.. i guess that was beta and they updated it and named it f8.

I tried messin around with settings man, but however my chip wont get past bios on 4ghz on anything less then 1.45V , maybe i do something wrong, but i tried your settings which u posted, and my computer would not even show picture just vents started and shutdown, and its the same untill i put voltage on 1.42+, but on that voltage it wont run 5min on stability test.. on 1.5V it runs prime95 for 8hours ...all ok. So i guess i got bad chip? But funny.. because on old motherboard p35c i could do 4.1ghz on 1.35V with the same chip. The batch is q813a168 which they say is pretty decent... maybe bad motherboard?

Another funny thing is , if i set 3.8 it runs on 1.28V perfectly stable.. My temps are always the same i got nice zalman cooler, ambient temperature is around 25C, and i got around 35C in realtemp idle and max 65 C on prime95 blend some hours..
 
Hello man! Thx for answers. Well about the bios.. i was checking gigabyte homepage for bios regulary every day.. 3 or 4 days ago there was suddenly f9d bios on gigabyte page, and it said 1333+ memory support, now 2 days ago i checked again and there wasnt any f9d anymore only new f8 with same description.. i guess that was beta and they updated it and named it f8.

I tried messin around with settings man, but however my chip wont get past bios on 4ghz on anything less then 1.45V , maybe i do something wrong, but i tried your settings which u posted, and my computer would not even show picture just vents started and shutdown, and its the same untill i put voltage on 1.42+, but on that voltage it wont run 5min on stability test.. on 1.5V it runs prime95 for 8hours ...all ok. So i guess i got bad chip? But funny.. because on old motherboard p35c i could do 4.1ghz on 1.35V with the same chip. The batch is q813a168 which they say is pretty decent... maybe bad motherboard?

Another funny thing is , if i set 3.8 it runs on 1.28V perfectly stable.. My temps are always the same i got nice zalman cooler, ambient temperature is around 25C, and i got around 35C in realtemp idle and max 65 C on prime95 blend some hours..

You need to flash your BIOS with F8. The same thing happened to me a month ago - Gigabyte had a seemingly newer version "F7C" up, I installed it, and it was pulled 2 days later and replaced with F7 which turned out to be way more stable. So first thing, flash with F8. Then try to get back to 4Ghz slowly - I couldn't go straight to 4Ghz after flashing with F8 either - crashes. I had to run 3.4Ghz, restart, 3.8, restart and then 4, slowly raising vCore to like ~1.3, ~1.35 and ~1.36.

If you're getting to 3.8Ghz with 1.28 Vcore then I don't understand why it would need 1.5V for 4Ghz. I don't know the answer. I'd flash F8 and try again, moving up slowly.
 
Hello,

I hope this won't be considered a thread hijack as I have the same motherboard and processor that Smittdog is using. I am new to overclocking and have always shied away from it until this new setup that I have:
Motherboard -- Gigabyte EP45-DS3L
CPU --E8400
RAM -- Mushkin 2Gx2 966599
Video -- Visiontek HD4870X2
PSU -- Corsair 650TX
CPU Fan -- Arctic Cooler Freezer 7 Pro
Vista Business 32bit OEM

I'm afraid I can't seem to overclock at all even after to copy the settings originally posted by Smittdog (although my version is conservative for now set to 3.6Ghz. When I tried to duplicate his settings, it won't boot and would revert to a fail-safe status):

These are the BIOS Changes I'm trying to implement:
Robust Graphics Booster -- Turbo
@ 3.6 Ghz (400x9)
PCI Express Frequency 100 Mhz
RAM @ 5-5-5-18
RAM performance switch [Extreme]
MCH Freq Latch 333 Mhz
System Memory Multiplier 2.40B
DRAM Voltage: -- 1.9V

CPU Vcore 1.375V
CPU Termination 1.2V
CPU PLL 1.55V

Everytime I run CPUz, it seems my machine is just running at stock speed.

Here are some screenshots I made--they may not be as relevant, but it might help you experienced OC'ers pinpoint my problems:

cpuz_ss.jpg

cpuz_mem.jpg

et6freq.jpg

et6volt.jpg


Any insight would be appreciated! And if I did hijack the thread, please let me know so that I may post at the appropriate forum.

Thanks much!
 
Flash to F8 and move up slowly (raise vcore and FSB ~300 Mhx at a time, rebooting each time). My PC won't boot either if I try to go straight to these settings from factory.
 
Thanks!
I have already installed F8 prior to posting. I went to Mushkin's forums and posted a similar question and they pointed out the proper RAM timing settings for me. Seemed to do the trick. I haven't gotten it to 4ghz like yours, but i'm comfortable with the stability of the 3.6ghz speed for now :)

Thank you
 
Hi all i have a very similar system to Smitdogg and tried using the above settings which didnt run stable for me. I'm pretty newb at this and I was wondering if anyone could advise me on which settings to tweak to get my system running stable at a reasonably increased overclock (if its even possible with my system).

I'm running:
Gigabyte GA-EP45-DS3L BIOS F7
standard Cooling
Vista 64-bit SP1
q9550
Asus 8800GT
2 x 2 gb 1066 mhz ddr2 corsair ram

Thank you for any help i can get.
got the same configuration here. awefully, I can´t get my corsair running @1066. could you or anybody else give me advice on the stettings? just a short post on how your settings look like would be sufficient. thx a lot guys!
 
got the same configuration here. awefully, I can´t get my corsair running @1066. could you or anybody else give me advice on the stettings? just a short post on how your settings look like would be sufficient. thx a lot guys!

Is your corsair like mine at 1066 or at 800 MHz and you are trying to OC it to 1066. Both might work if you are using expensive DDR2 modules (Ballistix).
 
Big thanx and to sum it all...

Big thanks to redduc900 for OC-ing advices, i spend so much time in front of the screen that sometimes i think i am too informed for advices and too numb for some new numbers, but you helped a lot (the fact u like Ducati only gives you more credit)!

What i learnt about EP45-DS3L and E8400:
They are great combo...

MAX tested CPU Vcore is 1.3875V
CPU Vcore forum suggestion is 1.275-1.325 (although most hits @ 1.375V)
My E8400@4GHz uses as little as 1.2750V to hit 4GHz
Intel specification Vcore 0.85V – 1.3625V
MAX temp 72.4°C, goes a little bit higher if u use intel stock cooler

To all Europeans i would advise buying Arctic Cooling 7 Pro cooler cause it`s a real bargain (17€) and keeps the CPU cool really good. Only thing to avoid is Arctic Cooling thermal grease, instead use STG-1 (Zalman i think).

CPU
CPU Vcore (default: 1.250V): use anything from 1.275 to 1.325V (think anything in this range is pretty safe). I got lucky, at 1.275V everything works @4GHZ. But if u plan to use OCed CPU only part time, for let`s say games or video encoding, like i will, u can ignore testing that is done by using Prime95 and lowering the upper voltage until u hit the stable frequency. Just use the lowest freq. that lets u boot and play the game. Starting at 1.275 is a good idea, and it won`t do any harm to your machine. It will give you the boost u need and afterwards u can go back to standard profile (i say profile cause i store BIOS profiles, one is OC profile, one is standard - one that i use when i don`t need speed, which is most of the time i spend online or listening to music...)

CPU Termination (default:1.200V): If u plan occasional boost leave it at default voltage but make sure you don`t leave it at auto or normal. Anything from 1.24 to 1.34 is good. Even extreme configurations use only 1.24V of CPU VTT, so no need to go overboard with this. I leave it at 1.200

CPU PLL (default:1.550V) Set to 1.550V, Auto over volts this way to much, like it does to the rest of the voltage hungry components. This board knows no limits, on Auto it will over volt simple DDR2 800MHz to 1.968 even when you don`t want it to and you didn`t overcloak it.

CPU Reference (default:0.805V): leave it on Auto

MCH/ICH

MCH Core (default:1.100V) anything from 1.2 to 1.4 goes, but @ 1.3V it`s pretty stable with 4GB of RAM. (1.24-1.28 likely good for 4GB, less is better if u want it to live longer). I use 2 GB so i use default voltage and it works.

MCH Reference (default:0.800V):Auto

ICH I/O (default:1.550V): 1.55-1.6, leave on 1.55 is my advice.

DRAM

DRAM Voltage (default: 1.800V): anything above default goes, but check your ram`s manufacturer web site and check the recommended voltage and timings. Can do a lot of good and can`t do any harm. If you don`t plan to overclock it, don`t leave it on auto, it over volts it, a lot.
 
More options explained

CPU Vcore
Normal CPU voltage.

CPU Termination (CPU VTT)
VTT is the termination voltage. It is also the voltage VREF is derived from. System Bus Input Supply Voltage, often referenced as FSB termination voltage (VFSB) simply the voltage the fsb is running on. VTT is the termination voltage for data lines used to interface the MCH with the CPU die(s) via the Front Side Bus. Higher values can provide additional FSB overclocking margin, especially with 45nm dual-core processors and quad-cores CPUs in general. We have never found any additional gains to be had above 1.36 when using air- or water-cooling. Setting Auto should default to 1.10V for 45nm CPUs and 1.20V for 65nm CPUs. The VTT helps stabilize the signaling between the NB and CPU cores. Never should be bigger than CPU Vcore.

CPU PLL Voltage
The PLL voltage is essentially the voltage applied to the CPU's internal clock distribution system. Selecting a higher PLL (phase lock-loop) voltage may help the installed CPU clock higher or may assist with maintaining stability when operating at higher FSB speeds. Most users will find they do not need to set this to anything above 1.50. Exercise caution when experimenting with higher values as there have been reported cases of CPUs losing cores after being subjected to voltages in excess of about 2.00V.

CPU Reference (CPU VREF)
Nominal GTL reference voltage for current generation processors/chipsets should be about 2/3 of VTT (or approximately 0.8v), although you may find your board's values a little lower (~0.75v). Intel's GTL specification dictates that each reference voltage should be tightly controlled as 67% of the current VTT voltage (0.67 times CPU VTT Voltage). Historical data has shown that dual-core CPUs (in particular 45nm dual-cores) often clock to higher FSB levels when the GTL reference values are closer to 63 ~ 64% of VTT while quad-core CPUs usually need the full strength value or even a little more voltage (67 ~ 70% of VTT).

MCH Core Voltage
MCH Core voltage is your Northbridge Voltage (Memory Controller) and the amount needed to be set depends on a few things. It depends on the FSB you are running, the Strap you are using, the actual Memory Multiplier you are using (Ram Speed) and the amount of physical Ram installed. P45 MCH is a bit tricky as well, sometimes when you think you need more less would be better and visa versa. All of course depending on the other mentioned items above.

MCH Reference
The NB (SPP) GTL reference voltage provides the same functionality as those for the CPU, the only difference being that the Northbridge uses only a single value. Tuning in the right GTL reference voltage (usually near the nominal 67% value) can sometimes lead to lower stable VDIMM requirements.

CPU clock drive/skew
In simple words, they alter the pattern of the current traveling to the CPU and PCIE bus respectively. They are only really used for extreme overclocks where high voltages don’t provide a regular current to the CPU or PCIE. At the blistering speed that today's components operate at, timing issues can rear their ugly heads when trying to communicate with each other and result in instability, particularly when overclocking. To keep the different parts working in sync, some motherboards -- such as this one -- allow you to introduce tiny delays in different subsystems of your PC. In this case, you're 'skewing' the speed of the CPU clock as measured in picoseconds (ps), or one trillionth of a second. Increasing from default 800mV to 900 can help to stabilize system after overclock.

MCH Frequency latch and SPD
Your north bridge has an internal clock speed and latencies just like your CPU and memory. The FSB of your north bridge can be found by dividing your original CPU multiplier by your set CPU multiplier and then multiplying by your FSB.
So if you are running a E6600 (266 * 9) at 400Mhz x 8 your NB FSB is:

(9 / 8) x 400 = 450Mhz FSB (1800Mhz Total)

Just like your memory may be able to run at 4-4-4-12 at 1000Mhz but needs to run at 5-5-5-15 at 1200Mhz, your north bridge has a series of latencies which it must adjust in order to maintain stability at its FSB. These latencies seem to play a far more significant role in system performance than memory latencies.
Intel has predefined specific latencies at specific NB FSB speeds. They are referred to as straps. There is a strap for when the NB FSB is 1066Mhz and under, 1333Mhz FSB and under, 1600Mhz FSB and under, ect. When you go from the 1066Mhz FSB strap to the 1333Mhz FSB strap, the north bridge's internal latencies loosen to allow for greater stability. For most of us the challenge will be to see if you can get on/stay on the 333MHz strap with an FSB over 400MHz and still have good timings and memory speed with a low tRD. Sometimes a very difficult balancing act to pull off, many times just not possible. In other words the trick is to stay on as low of a strap as you can but still have your FSB and memory where you want it.
Set on auto or at 400MHz to allow looser timings if you overcloak.

Static Read Control Delay (tRD)
When it comes to overclocking, the MCH functions as a hybrid of sorts. Like a CPU, it has an upper frequency limit and more voltage can often raise this limit. On the other hand, since it interfaces with memory it also behaves somewhat like memory with internal "timings" whose absolute values derive from the established FSB.

The outside world's first introduction to variable tRD settings came when a few overclockers noticed that setting lower MCH "straps" allowed for higher memory bandwidths. What they didn't know at the time was that they had unintentionally stumbled upon tRD. Tricking the motherboard into detecting an installed CPU as an 800 FSB (200MHz) part forced the MCH into setting a lower tRD value than if the FSB were 1066 (266MHz). Consequently, overclocking the system to the same higher FSB value with the lower strap setting yielded higher memory performance. Often times the effect was significant enough that real-world performance was higher even with a lower final FSB. The tradeoff was apparent however: a lower strap meant a lower maximum FSB. The MCH tRD value, just like a memory timing, must eventually be loosened in order to scale higher. What's more, as is the case with memory, additional voltage can sometimes allow the MCH to run with tighter "timings" at higher speeds.

Eventually the inevitable next step in memory performance tuning became a reality. The option to adjust tRD independent of MCH strap selection became part of every overclocker's arsenal. Nowadays the MCH strap setting does little more than determine which memory multiplier ratios are available for use. Although tRD adjustments are now possible in many BIOS implementations, some motherboard manufactures choose to obfuscate their true nature by giving the setting confusing, proprietary names like "Transaction Booster" and the like. Don't let these names fool you; in the end they all do the same thing: manipulate tRD.



Advanced literature
http://www.thetechrepository.com/showthread.php?t=87
http://edgeofstability.com/articles/dfi_p35/gtl/gtl1.html
 
Back