The Unofficial ASUS P5N-E SLI 650i Board Thread

[JustaGeek]

Originally posted by: Tuvoc
I tried to set my memory at 1000 5-5-5-15 and got the dreaded long beep meaning a memory error. Of course then the BIOS defaulted back to 800, which is in the "hole" area and it wouldn't boot at all again until I removed 1 stick of memory. It then booted I set it at 711, put the 4th stick back in, and all is well. That's the trouble with the 4 x 1GB memory hole, when you're experimenting it is a total pain to have to remove and refit the 4th stick when you hit a problem. I guess 5-5-5-18 or a little more viltage may have helped.

Anyway - running at 853 at the moment. When I increase memory speed I see a *tiny* increase in Sandra bandwidth, like 1-2%. Must be the board doing something with the sub-timings ? My numbers in outright terms are quite low - 4-4-4-12 2T at 711 was 5445 and at 853 is 5571. So a 20% increase in memory speed led to a 2% increase in bandwidth. And of course this is why my real applications see no benefit from the 2%.... I may as well go back to 711 and remain in spec. Could be this behaviour is specific to the quad cores

That's how I feel about overclocking the memory. You need it ONLY to accomodate the increased data bandwidth due to overclocking the processor. Otherwise, the risk of instability is too great, considering relatively minimal gains in performance.

I think you should leave it at 853 though, that should give it a sufficient throughput for your potential future overclock.

 

[Shampoo]

I've just about given up on getting my 4GB of ram stable.

I passed memtest and orthos cpu tests separately, but when I run a blend stress test in orthos it fails after 30 something minutes.

So strange.

Guess it's the memory controller on the board?

Going to try the new bios and see what happens.

I'm down to 900MHz trying it out

 

[Tuvoc]

It's not necessarily the memory - I guess just keep slowing it down and adjust the timings and see what happens - could even be a heat issue

 

[JustaGeek]

I found an interesting thread on OCZ memory forum:

http://www.ocztechnologyforum....showthread.php?t=27196

They recommend increasing the NB voltage to 1.5V with some memory!

 

[jeffreydeng]

I am having problem with running Vista. Below is my configuration.

C2D 4300 @3000MHz
OCZ OCZ2P800R22GK 1Gx2 pair @890MHz in black slot
EVGA Nvidia 7100GS GPU
Stock cooler

The problem I have is that Vista refuses to go to sleep mode. When sleep mode is forced, Vista came back immediately. It looks like that something is prevent the system entering the sleep mode.

This happens even I don't overclock the CPU and memory. I have reinstalled Vista several times and can not get rid of it.

Does any one have the same experience?

 

[JustaGeek]

You need to enable ACPI Standby state in BIOS to S1 and S3 - I am not sure if S3 alone is enough, but you can try. You can read about it in the Intel article here:

http://www.intel.com/support/m...sktop/sb/cs-020837.htm

 

[jeffreydeng]

Originally posted by: JustaGeek
You need to enable ACPI Standby state in BIOS to S1 and S3 - I am not sure if S3 alone is enough, but you can try. You can read about it in the Intel article here:

http://www.intel.com/support/m...sktop/sb/cs-020837.htm

Yes, the ACPI Standby state in BIOS is set to [S1&S3]. Any other ideas?

 

[JustaGeek]

Perhaps your Power Supply does not provide the current necessary for the S3.

Otherwise it is probably Vista related. Go to Power Options, and Select a Power Plan and/or adjust the other options in "Change when computer sleeps" window.

I'm not sure what else, my HP laptop always goes into S1 and then S3 when on the battery power; it does not go sleep by itself at all when the power is plugged in though.

I can, however, do it manually.

 

[jeffreydeng]

You are rigth that Vista has many problems with S3. I have three PCs at home. One is a Toshiba Laptop and it behaves the same way as your HP. It doesn't go to S3 automatically although it is set to. My second PC is a C2D E4300 with EVGA 680i lite mobo, I didn't have any problem with it regarding S3 mode.

Only this ASUS/E4300 combo has problem. When I press power button or do start>sleep, you can see that it tries to goto S3. The screen is off for 2 to 3 seconds and HDD is busy in storing, but it comes back with login menu. It looks like that some system device is preventing it from entering sleep. I don't have any other device and I can only think it is either the mobo or the 7100GS driver. I don't have a second PCI-E video card so I can not verify it. I am getting a 8500GT tomorrow and I will try.

Another interesting thing is that this PC doesn;t want to go hibernate either. When I force it to hibernate mode, it will do that and you can see that the system is powered off. But immediately it will resume itself. I am sure that these two are connected. But I just don't know what it is. I have been working on it for about three days and haven't found resolution yet. I visited lots of forums and lots of different cases are reported.

 

[JustaGeek]

Try unchecking all the devices' "Allow this device to resume from hibernate" or something to this effect in the Device Manager (you can wake it with the power switch until you find out what causes problems). Especially the Network Adapter - it might be the device that wakes it up. I am not sure how to do it in Vista - I know that it is an option in XP - the window that says "Allow the computer to turn off device to save power", and the other one is "Allow to wake up from Hibernate".

I hope you do not use the NVidia "network Manager" - if you do, uninstall it! It comes with the Chipset Driver package - you need the driver, you don't need the "Manager". It causes a lot of problems.

Check the wiring and make sure that there is no shorts in the wiring, or "leftover" standouts under the board. Do you have the 24 pin main board Power Supply plug...?

Make sure you have allowed hibernation in the Power Settings, and check how much is allocated for the Swap File (should be 1.5 times the installed memory).

Or... just leave it for a couple of weeks and see what happens, you can turn it off manually until you find out, right...? And check the Event Viewer for the System Error messages.

Also, see if your mouse/keyboard driver does not have any problems/conflicts - I have recently experienced problems with Logitech SetPoint v. 4.000....

Good luck!

 

[ElAguila]

Does the 0608 bios have the problem the the multi being locked?

 

[jeffreydeng]

Originally posted by: ElAguila
Does the 0608 bios have the problem the the multi being locked?

WHat do yo mean by 'the multi being locked'? I am using 0608 now and am not experiencing any issue except not being able to enter S3.

 

[JustaGeek]

Originally posted by: ElAguila
Does the 0608 bios have the problem the the multi being locked?

Apparently the CPU multiplier is unlocked in 0608, for the first time since 0401.

 

[JustaGeek]

Originally posted by: jeffreydeng
<div class="FTQUOTE"><begin quote>Originally posted by: ElAguila
Does the 0608 bios have the problem the the multi being locked?</end quote></div>

WHat do yo mean by 'the multi being locked'? I am using 0608 now and am not experiencing any issue except not being able to enter S3.

Your S3 problem is related to either software, or the hardware conflicts.

I highly doubt it is the issue with the board, IMHO.

 

[Noema]

Well, for the life of me, I can't get the darned thing to POST if I push the fbs over 1333. At 2.66GHz my E6400 is rock stable, RAM running @800MHz with relatively loose timings (5-5-5-15 - Stock) at 1.92v, but anything higher than that just fizzles...mobo won't even beep. I've tried upping the Vcore to even 1.4v (something I'm not to eager to do since I'm still in stock cooling, but I don't want to invest on better cooling until I can't make sure I can get it to POST at higher clock speeds); I've also raised the volts on NB with to no avail.

I was hoping I could get at least 2.98GHz but so far, no luck. Should I try the new 608 BIOS? I'm running 401 now, but neither 505 nor 602 helped.

Any suggestions? More Volts? Change multiplier? Or I just got a 'bad' board (I can live with 2.66 on my E6400).

 

[jeffreydeng]

Thanks JustaGeek for the useful advice in investigating issue related with S3 mode. I am still trying to figure out why but I am using it right now. Except to that my PC is up running. But the overclocking process is not normal. I can almost call it mystery. Here is my story

I have been struggling with this mobo and CPU set since last Monday when I received it. And I did many combinations of over clocking configurations and I found that it only works on following combination. (ALthough it could run at 1450MHz at 1.50v I am not going to do that until I get a decent cooler).

VCore: 1.45v
CPU clock: 1333MHz

Memory Volt: 2.178v
Memory clock: 890MHz

I am using auto for memory timing setting and CPUZ is reporting 5-6-6-31-25.

By the way, I am using OCZ OCZ2P800R22GK(OCZ DDR2 PC2-6400 Platinum Revision 2 Dual Channel) 1Gx2 pair memory.

I am using E4300 stock cooler and the SpeedFan reports 36C idle 46C full load. I am running Orthos which takes 100% of CPU time. The core template is 40C idle and 65-70C full load.

Before I decided to try this combination, it only worked at default clock (800MHZ for both CPU and memory). I tried 1200 MHz for CPU it even doesn't post. Another strange thing is that when I set CPU at 1200Mhz and memory at 800, when I try to change CPU clock from 1.35v to something else the BIOS stop responding.

I am using BIOS 0608.

This is really strange. I almost give up overclocking and tried to RMA the board or the CPU. But now I am happy. It looks like that this board has a overclocking hole between 800 and 1200MHz or higher which I haven't tried at all.

Has anyone experienced such strange behavior?

 

[jeffreydeng]

Originally posted by: JustaGeek
Try unchecking all the devices' "Allow this device to resume from hibernate" or something to this effect in the Device Manager (you can wake it with the power switch until you find out what causes problems). Especially the Network Adapter - it might be the device that wakes it up. I am not sure how to do it in Vista - I know that it is an option in XP - the window that says "Allow the computer to turn off device to save power", and the other one is "Allow to wake up from Hibernate".

I hope you do not use the NVidia "network Manager" - if you do, uninstall it! It comes with the Chipset Driver package - you need the driver, you don't need the "Manager". It causes a lot of problems.

Check the wiring and make sure that there is no shorts in the wiring, or "leftover" standouts under the board. Do you have the 24 pin main board Power Supply plug...?

Make sure you have allowed hibernation in the Power Settings, and check how much is allocated for the Swap File (should be 1.5 times the installed memory).

Or... just leave it for a couple of weeks and see what happens, you can turn it off manually until you find out, right...? And check the Event Viewer for the System Error messages.

Also, see if your mouse/keyboard driver does not have any problems/conflicts - I have recently experienced problems with Logitech SetPoint v. 4.000....

Good luck!

OK, here is the status of my PC now.

In device manager, for a handful devices there is POWER MANAGEMENT tab in device property. THese devices in my configuration are listed below.

- MOuse
- Keyboard
- Network devices (I have two, one is the onboard NIC and the other is the NetGear SuperG wifi adapter)
- USB root Hub ( I believe that these are from Nvidia drivers)
- HIgh Definition Audio Controller in SYstem devices (This is strange)

In POWER Management tab, there are two choices, one is 'Allow the computer to turn off this device to save power'. Other other is 'Allow this device to wake the computer'.

I have turned on 'Allow the computer to turn off this device to save power' for all devices and turned off 'Allow this device to wake the computer' for all devices.

The strange thing is that by default 'Allow this device to wake the computer' is turned on for High Definition Audio Controller device. This is really really strange. I don't know how this audio device can tuen on the computer. I am sure this is a bug whichis probably in either MS or Nvidia driver.

After I did above, I can force the PC to S3 mode. But I can not wake it up. There is no display at all although the fana are running. I have to power it down and up again.

But this is one step forward since I can hiberbate the PC now so that it takes less time to wake from S3 comparing with S5 mode.

So from now I am going to use Hibernate instead of Sleep and continuato figure out what is the cause of black screen when resuming from sleep.

 

[JustaGeek]

jeffreydeng,

I think your CPU is too hot under load - the max safe temp for Cores 1/2 shoud be 60C in my opinion. Check the processor's HSF brackets, if you have to, take a look on the back side of the MB to see if the pins are engaged properly.

And re-seat it on the CPU with a good quality paste - take a look at Arctic Silver 5 instructions at their website. You only need to place a line between the cores in the middle, although I personally just applied it all over the metal plate.

And very important - put a heatsink on the Southbridge! Some people don't have any problems, but mine was as hot as 62C without it.
I even put a little fan on it...

Regarding S3 - are you sure that your Power Supply supports it...? Do all of the 24 pins on the connector work fine (you don't need the - 5V pin - it is for the old ISA bus). You do use a 24 pin connector, right...? (Not 20 pin I hope).

Honestly, I would take it all apart and re-insert all the components back again. I had to remove my Audio Card to re-seat the USB cable, and had trouble re-inserting it - pushed it at an angle and it would just not go in! All the components shoud have a relatively tight fit, but you should never have to use any force to slide them in.

Hope this helps.

P.S. Just read your post again - I think it might be that Netgear Wi-Fi adapter. Take it out, use only the Ethernet to connect to the Internet, and see what happens... Or just disable one of them in the Device Manager...

I forgot one important thing: You MUST set the memory voltage, frequency and timings MANUALLY in BIOS. Choose Manual Overclocking, then (I believe for E4300) 800Mhz, then memory to 667 (good start!) or 800MHz, and timings with the 2T Command Rate.

 

[jeffreydeng]

The core temp I listed here were read from COreTemp which is always 10-15C higher than what is reported by SpeedFan which reports 50-55C at load. GIven this, is my configuration reasonable? CoreTemp reports core temp at near 80C for my laptop which uses Intel Core Duo CPU which runs at 2.16G.

As regarding S3, I really believe that it is because of driver of one component. I tried taking wifi adapter out and it didn't change. I am using all manual setting for overclocking and the PSU has 24 Pin connector as the main power connector.

I read too many complain related with S3 mode in Vista and realize that lots of third party drivers may not work with S3 mode.

By the way, I have another PC at home which is EVGA 680i Lite based mobo with E4300. I am using 7600GT with that PC and that PC is entering S3 mode fine.

I will report back when I figure out the reason for this.

THanks JustaGeek for all your help.

 

[Tuvoc]

I just updated to 0608 from 0505. Totally painless, it booted right up after the flash. I checked all my settings and they were all unchanged. Didn't bother to clear the CMOS as I've done before with this board. No other board I have had has required it.... Clearing CMOS for me means that the memory defaults back to 800, right into the 4 x 1GB hole, and I have to remove a memory stick to change the settings before re-fitting again...

 

[Shampoo]

I know people always say that you must clear the cmos after a bios update, but you don't really have to unless you realize any strange behaviour from the board.

I upgraded to 0608 and things are just fine and dandy.

I'm pretty much stable now.

Settings are as follows.

3.5GHz (FSB 437.5MHz x 4 = 1750MHz), 875MHz memory @ 5-5-5-15, 2T.
FSB:MEM, 1:1.
Going to try lowering my memory timings.

I'm 20 hours into a Blend stress test with orthos.

I know people use small ffts to stress the cpu, but to check overall stability DO stress with Large in place ffts or a blend test.

 

[JustaGeek]

Tuvoc, Shampoo,

I think that the "panic" about the CMOS reset after BIOS updates occurred due to the fact that people were not able to boot after the BIOS update, with the BIOS reading the EPP and the 1T and so on.

I have updated the BIOS on my 775Dual-VSTA numerous times, in Windows, turning off the Anti Virus and leaving only the explorer.exe and taskmgr.exe as the User running tasks (leaving all the System, Network Service and Local Service tasks running, too).

After the update, it was very important to enter the BIOS, "Load the Setup Defaults", "Exit and Save", and then re-enter the BIOS and do the rest of the manual setup.

Without loading the setup defaults after the update, some people would lose the Mac address, rendering the on-board Network Adapter inoperable and useless (you can't change the Mac address yourself, the board would have to be returned and re-programmed).

Glad to hear that from you, guys. When I decide that it is time to update (still running 0202. can you believe it...?), I won't lose my sleep over it LOL.

 

[JustaGeek]

jeffreydeng,

I would try turning off 'Allow the computer to turn off this device to save power' - perhaps they stay off in hibernation, and that's why you can't wake it up.

Otherwise, if the machine is working fine, leave it alone, you will figure it out later.

I think that CoreTemp does not read the E4300 temps properly, and you are right, there might be the 15C offset. If you don't experience any heat related issues, it is probably OK.

 

[Shampoo]

Hey JustaGeek.

You're still on 0202~!!???

Haha. I guess as long as you aren't running into issues you don't have to update your bios.

I don't like how 0608 takes longer to post than 0505 and 0401, but it seems to be doing well so far.

Better memory compatibility is part of the update for 0608, which is why I chose to update to it.

My guess is that the Asus techs probably relaxed the memory subtimings, since this board has been reported to use very tight memory subtimings by various review sites.

If that's the case, then things should be good with 0608 since memory subtimings don't have that much of an effect on performance, but more on stability in my experiences.

I did some major testing with my old 939 DFI Ultra-D board with memory timings. It had so many memory subtimings it wasn't even funny, but in the end they didn't do anything other than stabilize the ram.

Any reason why you're only at 325MHz FSB speeds JustaGeek?

Oh and by the way I am running 1.5v on my northbridge and everything is just fine.

Cheers,
Mike

 

[JustaGeek]

LOL I know, isn't that something... 0202... I feel like if I've been running and old machine... LOL

But... give me one reason to update! Absolutely no issues whatsover... so far...

And I really do not feel like being ASUS' guinea pig LOL.

Well, 325MHz... Again, stock HSF, perfectly stable machine, temps ~40C MB/CPU, ~46C Core 1/2 Idle, ~58C Load. IMHO, the safe upper limit for the CPU.

And my 3DMark06 score at ~5920 is limited by the 7950GT, but all the games play amazingly well at 1680x1050, on High or Ultra Quality settings. Quake 4 looks absolutely stunning at Ultra, and my Maple Leafs always win the Stanley Cup in NHL07 LOL.

Hope it comes true one day though...