ASUS P5N-E SLI 650i Thread

[Armalite]

Originally Posted by marsey99

nice post armalite, makes it easy to help

seeing as you have not overclocked your ram small ftt will be all you need to run to test your cpu.

tbh i think you should have no trouble running on those clocks with that cooling but im not sure you will get much more and stay stable.

1.4v is about as far as the stock intel blower can handle.

Thanks for the reply mate

I've hit a problem though, after running orthos, I fired up COD4 multiplayer, for only a single round, saw that my fps had gone DOWN in some spots on the more fps intensive maps (heavily grassed ones) so I googled the problem and apparently its from an unstable CPU Overclock, and a solution was upping the NB PCI-E frequency may help it out.

So I upped it from 100-105, and sure enough, it seems to have worked...in COD4 at least, I then fired up Crysis (demo) just to see if any improvement has resulted from the overclock, and sure enough my fps was a little better. Problem is though, about 3 quarters into the level, the textures on the rocks that are all about the level lost their texture, and just turned into a gray blobs with no textures which I've read is from the data being corrupted because of an unstable OC.

Should I up the PCI-E frequency to 110? or is this from the CPU not getting enough Voltage?

Any help would be appreciated

[marsey99]

you might be able to get abit higher than 110 from it, i think the most i got was about 117 on 1.39nb.

when i got gpu errors most of them were fixxed by drivers but adding vcore can help if its your cpu failing the gpu, if you get me.

just keep an eye on your temps till your happy with them.

[Armalite]

Originally Posted by marsey99

you might be able to get abit higher than 110 from it, i think the most i got was about 117 on 1.39nb.

when i got gpu errors most of them were fixxed by drivers but adding vcore can help if its your cpu failing the gpu, if you get me.

just keep an eye on your temps till your happy with them.

Thanks again

Should I change the NB voltage? all voltages are set to auto, except the CPU Vcore which is at 1.35000v. I had no idea you had to overclock the PCI-E bus and stuff like that. I've read sooo many different responses, some say keep it at 100, some say 105-110, and some say 115+. Alot of people think taking it over 100 will blow/fry the GPU?

Drivers I'm using now (175.16) are not up to date, but are the best drivers I've ever had with this card ( 8600GT ) 175.19 really mess up my fps in games, so I reverted back to 175.16.

Here I was thinking I was getting close to my first (successful) overclock. This is definitely below the belt, am I in over my head here?

[marsey99]

try 177.35, they are what used in the 3dmark 06 run in my sig.

it all wants to oc once you start to mess with the fsb, the trick is to ballence it all out.

can you see what it says on the chips of your ram, or what does cpuz call them?

it is most likley autoing your nb to 1.2 or 1.39v, you could try both and see if it effects your temps any. lower temps are allways good on this board.

[Armalite]

Originally Posted by marsey99

try 177.35, they are what used in the 3dmark 06 run in my sig.

it all wants to oc once you start to mess with the fsb, the trick is to ballence it all out.

can you see what it says on the chips of your ram, or what does cpuz call them?

it is most likley autoing your nb to 1.2 or 1.39v, you could try both and see if it effects your temps any. lower temps are allways good on this board.

Thanks mate, I'll give those drivers a whirl.


Lol, In the SPD tab in CPU-Z it is blank in the Manufacturer label. On the chips themselves, it has a code: DDR26671GB-RVM6411010
Thats all it says on the ram sticks. So I guess its cheap stuff.

Will try the voltages later.

EDIT: Ok, set the NB voltage to 1.2, Both cores idling between 24-26, although it is early morning and it being winter helps. In CPU-Z in the 'SPD' menu, under timings table, under 333 MHz it says
5.0
5
5
15
20
But under the 'Memory' menu it says the ram is running at
1:1
5.0
5
5
31
20
2T
should that 31 be a 15? Should I enter the 333MHz timings into the BIOS manually?

cheers

[Gauvenator]

@ armalite: at 333fsb you should only need the northbridge to be on auto voltage, as it is an officially compatible speed. Your cpu is what needs the extra voltage. Imo, don't mess with the pcie frequency, stick with upping the cpu vcore.
Your instability could also be caused by that loose memory timing, so make sure to manually set the timings in the bios.

[Armalite]

Originally Posted by Gauvenator

@ armalite: at 333fsb you should only need the northbridge to be on auto voltage, as it is an officially compatible speed. Your cpu is what needs the extra voltage. Imo, don't mess with the pcie frequency, stick with upping the cpu vcore.
Your instability could also be caused by that loose memory timing, so make sure to manually set the timings in the bios.

Thanks for the reply.

Ok, so I just put the NB voltage back to auto, set the PCI-E frequency back to 100, upped the CPU voltage by 1 step from 1.35000v to 1.35625v, set these timings CAS# Latency to 5, RAS# to CAS# to 5, RAS# Precharge to 5, tRAS to 15. Was I meant to set the tRC to 20? After checking CPU-Z in the memory tab it now says
1:1
5.0
5
5
15
20
2T
as stated above, this is what the timings table says it should be at 333mhz, so I'm guessing the Ram is now running in correct timings. CPU is idling at 28\27, running a small FFT test in orthos, which is at the 17 minute mark right now, CPU load is at 56-57, I'm pretty sure this is really pushing it now? if it doesn't go up any further, will I be ok? I've read 60c is the max you wanna go with a C2D.

Thanks for your help

[Gauvenator]

Originally Posted by Armalite

Thanks for the reply.

Ok, so I just put the NB voltage back to auto, set the PCI-E frequency back to 100, upped the CPU voltage by 1 step from 1.35000v to 1.35625v, set these timings CAS# Latency to 5, RAS# to CAS# to 5, RAS# Precharge to 5, tRAS to 15. Was I meant to set the tRC to 20? After checking CPU-Z in the memory tab it now says
1:1
5.0
5
5
15
20
2T
as stated above, this is what the timings table says it should be at 333mhz, so I'm guessing the Ram is now running in correct timings. CPU is idling at 2827, running a small FFT test in orthos, which is at the 17 minute mark right now, CPU load is at 56-57, I'm pretty sure this is really pushing it now? if it doesn't go up any further, will I be ok? I've read 60c is the max you wanna go with a C2D.

Thanks for your help

Your c2d is a G0 so you're good up to 72°C according to http://processorfinder.intel.com/det...px?sSpec=SLA9X
If orthos lasts for a few hours your oc is pretty stable, but 8 should rule out most instability.

[deltaepsylon]

Hehe, quick random question which I think pertains to the board/bios. How can I disable cores on my QX6850? I've looked through the bios everywhere, cant seem to find it but I recall hearing that the disabled cores settings were in the bios...helpz?

-Thanks

[Armalite]

Originally Posted by Gauvenator

Your c2d is a G0 so you're good up to 72°C according to http://processorfinder.intel.com/det...px?sSpec=SLA9X
If orthos lasts for a few hours your oc is pretty stable, but 8 should rule out most instability.

My E6600 is actually a B2 stepping 6. I know G0 are the best overclockers, so yer at about 1 hour 30 minutes it did end up hitting 60c so I stopped it, so I'll be looking into buying a ThermalRight 120 Ultra eXtreme, but for the time being, I'm fairly certain its stable, its running alot more smoother after applying the settings you recommended. idling at 30c, and in games its gets in between 44-48 max, during midday in Australia. Tonight when the temperatures go down to about 9-12c I'll run an overnight Orthos test for stability.

Thanks again guys for the help, Appreciate it