I am only 5% OC'ed and already black screen of death???

A

audiobahn1000

Member
The equipment I have is:

CPU: Core Two Duo 2.13 GHz. E6400
RAM: 2 GB. Corsair 800 MHz. XMS2 DDR2 (dual channel configuration)
GPU: X1900XT 512MB Connect3D
Mobo: Gigabyte GA-965P-DS3
PSU: 535 watt Enermax Whisper II
HD: 250 GB. Seagate Barracuda 16 MB.
OD: 52X DVD±RW
Monitor: Hanns-G 19" DVI LCD widescreen



I started overclocking the CPU in the BIOS tonight. I got to only 2.232 GHz. and I couldn’t boot. However I know that if I leave the computer unplugged for a few seconds the mobo woill disable the overclocking and let me reboot. So I did that and entered the BIOS as soon as booted up and reenabled the overclocking and Windows started fine overclocked at 2.232 GHz. Then after I ran all my benchmarking tests and such I restarted again to overclock more. This time I choose 2.264 Ghz. I saved it in the BIOS and once again I could not boot. So I unplugged the computer again to revert to OEM settings and rebooted into the BIOS. And once again I reenabled the overclocking but this time I increased the vcore from 1.325 volts (stock voltage) to 1.33125 volts (one click up) and saved and exit. The computer restarted and once again I could not boot. So I unplugged the computer again and reverted to OEM settings and I just left the overclock off and that is where I am now.

So what’s going on? I talked to people with this same processor that said they did not even have to touch the vcore until 2.8GHz. I am but barely 5% OC’ed.

Maybe I am doing something wrong. This is how I went about things:

I started out at 2.175 Ghz. I ran PC Wizard 2006 and benchmarked the CPU. I wrote down the Whetstone (FPU) Wheitstone (SSE2) and Dhrystone results. I then keept the program open to monitor CPU temp, voltage and loading.

Then I ran two instances of Prime 95 (one for each CPU) and ran the torture test on each and choose the in-place large FFT’s for max heat consumption. I ran the program for about 4 minuets with both CPU’s loaded 100% and recorded the average CPU temperature.

Then I close Prime 95 and open CPU Stability Test 6.0 and test the computer for stability for about 4 minuets.

It was stable every time.

As far as the steps I took in overclocking and the temperature recorded I got:

1: 2.175 Ghz., 47-48 C
2: 2.207 Ghz., 48-50 C
3: 2.232 GHz., 49-50 C
4: 2.264 GHz., 50-51 C

So whats wrong?
 
That's pretty warm, what kind of cooling do you have?

EDIT:For comparison I have a E6300 at 2.3ghz, and its only 37c Full load with Orthos SP2004
 
Last edited:
spacedude89 said:
That's pretty warm, what kind of cooling do you have?

EDIT:For comparison I have a E6300 at 2.3ghz, and its only 37c Full load with Orthos SP2004
Click to expand...

I have stock cooling. HOWEVER! It only gets that how when I run dual instances of Prime 95 with the max heat setting choosen. At idle its only like 37 C warmed up. And when fully loaded in my stability testing program its only like 41 C.
 
That is still pretty warm, Idle I'm at 28-29c. I suggest getting a better heatsync, That may or may not be your problem, but it is still very advisable if you plan on overclocking.
 
50ºC under full load is okay...but 50C idle is too hot.

So if your temps were 50C while running benchmarking and stability programs (or another program that uses 100% of your CPU), that's about normal...shouldn't be causing your system to be unstable.


Maybe there's another component being overclocked unintentionally...are your PCI ports locked?

EDIT: 37C idle is fine....a bit on the warm side, but still fine.
 
computermaineack said:
50ºC under full load is okay...but 50C idle is too hot.

So if your temps were 50C while running benchmarking and stability programs (or another program that uses 100% of your CPU), that's about normal...shouldn't be causing your system to be unstable.


Maybe there's another component being overclocked unintentionally...are your PCI ports locked?

EDIT: 37C idle is fine....a bit on the warm side, but still fine.
Click to expand...


It is not 37 C at idle. Its 37 C at idle right after I close Prime 95. When I first start the computer its only like 27 C.

I do not know if PCI-e is locked. How do I check? And how does that have anything to do with me overclocking just the CPU and RAM?
 
audiobahn1000 said:
It is not 37 C at idle. Its 37 C at idle right after I close Prime 95. When I first start the computer its only like 27 C.

I do not know if PCI-e is locked. How do I check? And how does that have anything to do with me overclocking just the CPU and RAM?
Click to expand...

Your temps are fine...

The PCI-e can be overclocked unintentionally while you overclock other components.

Praetor said:
It is also important to note that, sometimes, overclocking the core clock will also inadvertantly overclock other things like the AGP or PCI bus. Overclocking friendly boards tend to have 'things' called "locks" that essentially force those devices like the AGP and PCI bus to run at their standard speed yet still running the remainder of the system at a faster speed. Not all motherboards have these locks though and as such you may be running into OC ceiling as low as 10MHz into an overclock ... do some research on your motherboard model to find out. Also, not all motherboards that have such locks have functional locks: a good example of this is the ASUS A8V Deluxe, Rev 1.xx ... the locks were supposedly in place but it's well known that they really arent functional and should be treated as it they were not actually there.
Click to expand...
 
Dammit this problem is anoying. I got it to boot OC'ed to 2.264 again but then when I restarted it again I get nothing. wtf. I shouldent be haveing this problem untill like 2.8 GHz.
 
you can also check out this, from the same post...I take it you haven't read the OC101...

Praetor said:
I only OC'd by 30MHz on the core but It doesnt POST!! Does this mean I've hit my ceiling?
Most likely not. This is where the power supply comes in and starts to play as a factor on your OCs. The general idea is that, the harder you push your system, the more power it requires (shouldn't be a surprise there) and as such what this means is that you will have to start increasing the voltage being delivered to your CPU (designated as Vcore).

A good method of approaching this is to OC until you cant boot (this is why you approached it a 5-10MHz at time) anymore and then increasing the Vcore by a single step (usually 0.025v). If that wont boot still, increase it by another step. If three increases on the Vcore still dont allow you to boot, it becomes more plausible that you've reached your OC ceiling. Also note that in systems where, by OCing the core clock, you also increase the memory clock, you may also consider increasing the amount of voltage going to the RAM (designated Vdimm) by a single notch (usually 0.1v sometimes less). You shouldnt however, unless you know what you're doing, go past 2.8v on the voltage for DDR memory and 2.0v for DDR2.

It's also important to note that long term, it's not heat that kills a CPU but rather voltage ... this explains why some guy with a Vapochill unit keeping his box at -10C can still fry his CPU. When you start pushing the Vcore a lot (i.e., say by 0.4v+) you run the risk of long term damage. Same dealy goes for RAM.

When you OC always make sure you keep an eye on your temperatures and voltages to see that the temperatures dont get out of hand a your voltages stay relatively close to expected values. Three tools that come to mind are SpeedFan, MBM5 and for ASUS motherboard users, ASUS Probe. If you dont have an OS installed yet or anything, manual temperature probes as well as the BIOS hardware monitor are good places to start.
Click to expand...
 
im going to guess its the power supply.
you definitely dont need any overvoltage, because that raises heat a lot.
now, on the northbridge you may need some overvoltage because the fsb is running so high. its probably just your particular mobo, i have the same model (not the same northbridge) but the passive cooler is also a wall for the DS3 and other core 2 duo gigabyte models. i dont know why they did that... it cools it fine at stock settings but why they would add in all those extra o/cing features and then throw a chunk of metal with no fan on the northbridge.... i dono...
anyways you may want to upgrade that power supply since you're running an x1900 card.. and if they'll take your mobo back for an 85% refund that may be a better choice, i mean i overclocked mine 800mhz or so and i still think im not going as fast as i should.. im running it at 1.000V though lol
 
Last edited:
Its a 535 watt PSU. Even under full load the PSU does not draw more then 150 watts from the outlet. I measured it. How can 535 watts not be enough? Everyone told me this is a good PSU too befre I bought it?
 
I have the DS3 + E6300 @ 3.2 Ghz no problem. Are you overclocking the right things? You should overclock the MCH +0.1v :)
 
should be in oc101.

but basically everything runs of the base clock (in core 2 duo's case, 266mhz)
you raise it to 267 mhz it runs the cpu at (267*8) instead of (266*8) overclocking you by exactly 8mhz.
this also effects the memory because if it runs it at a default clock of around ddr2 800 (266*3) then lowering it to 2.5 rather than 3 would make it run on a divider (ddr2 667 = 266*2.5). some motherboards say it differently, and its referred to as a divider because in the past memory has been run at frequency's lower then the base clock... but now its ddr2 and were still calling it a divider so w/e
technically ddr2 800 is 400mhz though, but some motherboards call it different things..
 
Last edited:
You must log in or register to reply here.
Back
Top