Overclocking Escapades: Intel socket LGA775 Core 2 Duo E8400 E0 stepping

[Bobnova]

400FSB strap on P45 chipset is generally garbage. 333 is usually your best bet.
More NB volts may well be the answer. Definitely check the RAM voltage, timings, and speed.
What's your PLL set to?

What revision UD3P is that?


As to the VID/max volts question, VID is completely unrelated to maximum safe voltage. It is the range that the socket and VRM controller will allow a CPU's internally pulled LOW and HIGH VID pins to set the voltage to.
Hence, it's a per-socket (or even more specifically, per Intel VRM Spec) rating.

 

[(G{in}[AK)TION]]

-RANT MODE ENGAGED-

the dx48bt2 is frustrating! i was able to keep the processor at 3.6ghz, now i cant at all! i set all to 1.4v, nothing. started at the lowest and then moved up little by little with mch and then fsb, still nothing!

intel should have made their OWN boards with MORE quality in them instead of having foxconn do it! i dont care what was going on! they are intel and they have the funds to do it!

i may as well hope that i can get my hands on a gigabyte board! and if its asus, then i hope their older boards dont spontaneously grow defect and then die on me like their previous boards did!

EDIT: and lastly, i am stuck at 3.5ghz!

 

[EarthDog]

Welcome to Intel motherboards Gin... don't buy that crap and expect good overclocking.

This is also tech tweakers thread you are hijacking. You will want to start your own.

 

[Tech Tweaker]

400FSB strap on P45 chipset is generally garbage. 333 is usually your best bet.
More NB volts may well be the answer. Definitely check the RAM voltage, timings, and speed.
What's your PLL set to?

What revision UD3P is that?


As to the VID/max volts question, VID is completely unrelated to maximum safe voltage. It is the range that the socket and VRM controller will allow a CPU's internally pulled LOW and HIGH VID pins to set the voltage to.
Hence, it's a per-socket (or even more specifically, per Intel VRM Spec) rating.

My CPU Termination voltage (Northbridge) is at 1.4v, MCH Core is at 1.3v, RAM is at 2.2v. I haven't touched the CPU PLL, it's at the stock 1.5v.

It's revision 1.1.

RAM is at 900MHz with 450MHz FSB, 5-5-5-15 timings.

-RANT MODE ENGAGED-

the dx48bt2 is frustrating! i was able to keep the processor at 3.6ghz, now i cant at all! i set all to 1.4v, nothing. started at the lowest and then moved up little by little with mch and then fsb, still nothing!

intel should have made their OWN boards with MORE quality in them instead of having foxconn do it! i dont care what was going on! they are intel and they have the funds to do it!

i may as well hope that i can get my hands on a gigabyte board! and if its asus, then i hope their older boards dont spontaneously grow defect and then die on me like their previous boards did!

EDIT: and lastly, i am stuck at 3.5ghz!

Strange you're having so much trouble. I had difficulty with that same board too, but not as much as you seem to be having.

I never had to go over 1.3v on the MCH with that board.

 

[Bobnova]

Try 1.56 and 1.6PLL.

 

[doz]

MCH Core is your northbridge voltage, not the CPUTerm. I would leave CPUTerm on auto for the moment. Id set your NBv to 1.4v (MCH Core).

Also make sure load line calibration is enabled and just set 1.38v for the CPU voltage. That should easily get you to near 4ghz or above and give you enough play to ensure the CPU voltage is not an issue. If LLC was not enabled and you had a low CPU voltage, that might also cause an issue.

 

[deathman20]

Man been a while but I know my E8400 did 4Ghz no issues, eventually near the end before I upgraded to a 920 I was doing 4.2Ghz without a sweat.
The E8400 was a very very well done CPU with great overclocking potential.

 

[OptyTrooper]

The voltage jump from 4GHz to 5 is tremendous.

 

[Bobnova]

The voltage jump from 4GHz to 5 is tremendous.

+1 to that.
Often requires Cold as well as Massive Volts.

 

[OptyTrooper]

+1 to that.
Often requires Cold as well as Massive Volts.

My Cryo Z will allow me to hit 5Ghz @ 1.8v.

As an aside I found out it was pulling down fine. Turns out the thermometer was reading wrong I shot the socket temp with a laser and it read -43.

 

[Tech Tweaker]

This is most definitely an issue isolated to this motherboard as far as I can tell, as I doubt this is an issue with the P45 chipset.

I've increased MCH Core to 1.4v (also tried 1.42v, didn't want to go higher as I read that going above 1.4v with a 45nm CPU is risky (if not dangerous) to the CPU), CPU vCore at 1.4-1.45v, ICH Core to 1.3v (from 1.2v), CPU PLL to 1.57v (and 1.59v), and it still won't run stable past 3.6GHz (400MHz FSB). I notice I have a pin in the CPU socket which appears to be broken off, that may be related to, or causing my issues.

I put the same CPU, same cooler, same video card, same memory, and same power supply with an Abit IP35 Pro and I was running at 4.3GHz (475MHz x 9=4275Mhz) with no trouble at all. The system was completely stable as far as I could tell. CPU vCore was 1.4v in bios (though oddly it showed 1.28v in CPUID Hardware Monitor, don't know what that was about) and increased MCH Core on that board to 1.4v, and set memory timings to appropriate settings at 2.0v. That was all I did, and it ran flawlessly, I even downloaded and installed updates and installed programs with the system at those settings with no glitches or lockups.

 

[doz]

Have you attempted to find the boards sweet spot? Setting 400mhz does not always work. I had a picky P5Q-Pro and one that was not picky. It loved 445x9 but did not like doing 500x8. It would also have issues at certain mhz in between. Did you try 500x8? Some boards love that, others do not.

 

[Tech Tweaker]

Have you attempted to find the boards sweet spot? Setting 400mhz does not always work. I had a picky P5Q-Pro and one that was not picky. It loved 445x9 but did not like doing 500x8. It would also have issues at certain mhz in between. Did you try 500x8? Some boards love that, others do not.

As far as I can tell it doesn't have a sweet spot. I've found a lot of spots it doesn't like though. 412-450MHz is a no-man's land. 412MHz is slightly unstable and tends to crash when shutting down or restarting (and sometimes when firing up certain programs or running benchmarks), regardless of voltage settings (high voltage or low voltage doesn't matter (1.35-1.45v)). Anything beyond ~415Mhz tends to crash the system very shortly after entering the OS with default 9x multiplier, or before it even makes it into the OS. 445-450MHz is very unstable and it doesn't even make it into the OS usually, often it blue screens before it even attempts to load the OS.

The board seems to like anything at 400MHz or lower, and that's it.

Testing 500Mhz x8 now. It did at least make it into the OS at this setting. I'm not quite sure why lowering the multiplier and increasing the FSB would allow it to run more stable if I'm at the same effective core speed.

 

[doz]

FSB is more board limited. Different boards act differently. Why do you think motherboards were so important during P45 days? Everyone was looking for a cherry board that could do 600mhz FSB. They were few and far between.

 

[RJARRRPCGP]

Setting 400mhz does not always work.

I could see that if it's the P5QL Pro. (P43 version)

Is P43 a defective P45?

 

[Tech Tweaker]

Why do you think motherboards were so important during P45 days? Everyone was looking for a cherry board that could do 600mhz FSB. They were few and far between.

I honestly wouldn't know, I ran a socket 754 AMD rig from 2005-2011. So I missed the 775/P45 days entirely and didn't build with a 775 board (G33 chipset aka the P35's little brother) until 2011, didn't use a P45 board until 2013.

 

[doz]

I could see that if it's the P5QL Pro. (P43 version)

Is P43 a defective P45?

All boards are different. You would that even the better boards would be consistent but they are not.

I had one P5Q-Pro that would do 533mhz but not 500mhz if you can believe that. I also had a UD3P that would hit 476mhz but not 445mhz. All these numbers were with the same chip.

 

[DocClock aka MadClocker]

Try one stick of mem at a time, and sometimes just swapping places between the two dimms will make the difference My old sdram is like that

 

[Tech Tweaker]

Try one stick of mem at a time, and sometimes just swapping places between the two dimms will make the difference My old sdram is like that

Already tried. I tried single stick, tried each slot on the board. My memory just happens to be one of those kinds that likes a specific stick in the main DIMM slot (if I put the other stick in the first slot it won't boot most of the time unless I manually set the voltage, if I switch the two though it runs fine).

 

[Tech Tweaker]

I started testing this CPU again.

Seems my EP45-UD3P was the weak link last time, as my Evga 780i SLI is doing much better so far.