Windows 7 locks up...

[Cr00zng]

I am puzzled by the symptoms of the new system...

Specifications:

• CPU: i5-2500K
• CPU cooler: Noctua NH-C14
• Motherboard: Asus P8P67 LE
• OS drive: OCZ Vertex 3 120GBs
• HDD: 2 x Seagate Barracuda 7200.12 ST31000528AS 1TB 7200 RPM (SATA 3)
• Optical drive: Lite-On iHAS424-98 (SATA)
• Memory: Corsair DDR3 1333 Model CMX8GX3M2A1333C9
• Video card: HIS Silence H557HR1G Radeon HD 5570 1GB 128-bit DDR3 PCI Express 2.1 x16

The system runs well and has around 30 Celsius degree temperature at idle. With the Prime95 torture running for a half an hour, the temperature is 60 Celsius degree. The fan-less video card's temperature is around 45 Celsius degree. The motherboard temperature is couple of degrees lower than the CPU idle temperature. All the latest drivers, service packs, patches had been installed.

The issue is that the system locks up solid; no keyboard/mouse control is possible and the reset button does not work. The only way to restart the system is with the power switch. Upon reboot, Windows does complain about "improper shutdown"; however, no errors in the event logs.

The lock up could happen using any of the programs, Firefox, Macrium, etc., but it did not happen during the torture test with Prime95 using 98% memory utilization. I have a feeling that this is some hardware related problem, but I am lost as to where to start?

Does anyone have an idea where to start?
TIA....

 

[SlasherIT]

Check your RAM with Memtest.

 

[Cr00zng]

Well, Memtest+ ran for over two hours, did two paths, and came up with no errors. The Prime95 ran for over five hours and also came up with no errors:

Torture Test completed 435 tests in 5 hours, 34 minutes - 0 errors, 0 warnings.

I've rechecked Corsair' site and it still states that the memory works with the Asus motherboard. On the other hand, Asus currently does not list this memory (I would swore they did) at their site:

CMX8GX3M2A1333C9

They list this instead:

CMX8GX3M4A1333C9

The difference between the two model is the number of sticks; the former one that I have is 2x4 GBs, while the Asus listed one is 4x2 GBs. The other is the voltage with both having SPD based settings of 1.5V and the M4 has been tested at 1.65V.

Interestingly, Asus lists GSkill and OCZ as the two OEMs that have 2x4 GBs memory sticks that work with the P8P67LE board. Could the number of sticks and their sizes cause this kind of issues?
TIA...

 

[SlasherIT]

Hmmm... Do you have any other spare left over RAM you could use to test if the problem still occurs with a different RAM type?

If you do, remove the existing RAM you have and replace with the left over one's you have, and test if the issue still occurs.

 

[Cr00zng]

Unfortunately, I don't have any spare memory that would fit this box. The memory settings had been re-adjusted in the BIOS, or EFI with this board to:

1333/9-9-9-24 and 1.5 Volts.

The things different from auto detecting the memory is the voltage, 1.6 auto vs. 1.5 manual; doing so also removes the performance, or XMP profile for the memory.

This is evidently a know issue with the P8P67 LE board and Corsair's XMP type memory sticks. Here's one of the links for similar issues:

ASUSTeK Computer Inc.-Forum- Memory or Motherboard problem?

Interestingly, the guy who had the same issue as my machine does was actually using the Corsair's CMX8GX3M4A1333C9 with his P8P67 LE board, which is 4x2 GBs set or the Asus recommended four sticks. I guess it doesn't really matter, if one follows board makers recommendation to the letter...

It was sort of evident that there's some hardware issue since when started at the first time the machine took a double take. Meaning that it did start up, shutdown by itself without going through the BIOS, and started back up after about five seconds, and loaded the installation DVD. It still does it when the system is cold booted and rarely when it's rebooted.

Getting another set of memory is an option, but the issues make me think twice. Is it possible that this Asus board has some issues with relatively high density memory sticks? Sure, it certainly could be Corsair in this case but other 4 GBs sticks could exhibit similar symptoms as well. Maybe give Gskill a try, although when a company website has a memory finder application that does not work, it makes me kind of weary...

 

[Shootist]

Pull the heatsink and reset the CPU.
The board will do that ON/OFF/ON when the CPU can not be detected at first power on.
Mark sure you clean the socket and the bottom of the CPU with some Alcohol. Let it try completely, 5-10 minutes, before putting it back in.

 

[SlasherIT]

Hmmm, thanks for the info. Now I am thinking its the motherboard or CPU that is the main root of the problem. Try doing a complete CPU stress test on Prime95. Try leaving on for as long as possible. Post back the results if anything strange happens.

 

[Cr00zng]

I can reset the CPU after it's cleaned up and probably will do it too after couple of days. It just seems at the moment that changing the memory voltage did make a diffence; more testing needs to be done to be sure.

It seems that the "double boot" issue is Asus board specific and can be alleviated by changing the memory. Here's one vote for Gskill memory that doesn't trigger the double boot:

Newegg.com - G.SKILL ECO 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model F3-12800CL8D-8GBECO

Quote from the above link's feedback section:

They're also on the approved RAM list for the P8P67-M PRO which I suspect is why I don't get the 'double boot' issue a lot of folks have reported.

The longest I've ran Prime95 is five and a half hours, the system ran just fine with 98% memory utilization and 100% CPU (all cores) utilization. The CPU temperature was around 60 degree Celsius, which is not bad for a quite machine. It's about two feet from me and I can hardly hear it. Maybe I leave it one for overnight and check it in the morning...

 

[SlasherIT]

Ok, the CPU seems to be fine in that case. Its (thanks for clarifying) definitely between the motherboard and RAM. You should always check reviews and comments on anything you buy to make sure its all okay and no problems will arise. I suggest that instead of buying new RAM as your current RAM is pretty good, and it would be a waste of $125 to get new RAM, to take this on with Asus. Your motherboard comes with a 3 year warranty, so of course the guarantee is still in place. You should request from Asus to get another similar as possible motherboard, or ask to fix it by maybe creating a BIOS update that will fix the problem. Its your choice ultimately on what to do, but I suggest (unless you like to spend cash freely) not to buy separate RAM to fix this issue. Its better to fix than avert the problem

Slasher

 

[Cr00zng]

Well, to be frank, the specification for this system had been put together in January and put on hold, due to Intel's SATA 2 chip issue. There was nothing about certain memory may not work with this board at that time. And yes, I should've checked prior to ordering...

I'll take it up with Asus, but they may just send me packing. The current QVL does not list the Crucial model # CMX8GX3M2A1333C9 as compatible. Let's see where it gets me...

The board BIOS is at the latest version and it did come with already installed. Some people had success resolving this issue by clearing the BIOS with the jumper, while others with the same setup have not been successful. It could very well be some quality control issue with the early Revision 3.0 boards...

 

[SlasherIT]

Hello,

Wait, is your desktop custom built or from an OEM?

Ok good choice taking it to Asus, from what I hear, they have good customer service, so you should get it solved quickly (hopefully). You could try Reinstalling the BIOS, even though you have the latest version installed. You never know, maybe it could be fixed, maybe not. Its worth giving it a shot.

Slasher

 

[Cr00zng]

This is a custom built and yes, I either use Asus or MSI boards...

Here's what Asus says about this issue, quote:

Please clear CMOS for a try.
If no effort,I suggest you to take the motherboard out of case and keep the necessary components(a CPU, one memory, a video card, a HDD and a ODD) onboard to re-install a new clear OS and try again.

If still no help,please first try to have a further check with your vendor, if possible.

Well, clearing the CMOS and/or reinstalling the same version of BIOS are pretty much we've been discussing so far. Despite the fact that the temperature in this system is just fine, Asus still wants to run this board "naked". No thanks, I rather RMA the board at that point...

Interestingly, the system does not display double boot anymore, nor have I've seen it freezing during the last couple of days. Maybe it burned up some of manufacturing "grease" or something

 

[SlasherIT]

Hey,

So have you re-installed the BIOS, and then everything started working again?

Is the problem fixed?

Slasher

 

[Cr00zng]

Sorry Slasher, I've got distracted with things...

This is what had been done:

• changed the memory (yes, the Gskill)
• reset the CPU (mainly due to the size of the heatsink that prevented access to the first memory slot)
• clear the CMOS with the jumper

Neither of these resolved the random freeze issues and I was pretty much ready to return everything. After some searches, I found couple of people on the web who do experience similar issues and practically tried every one of them. It seems that the resolution for my machine had been a combination of fixes, but if I'd need to bet on it, I'd take the Intel resolution stated in my posting here. Maybe the video driver in this posting may had something to do with it as well.

The BIOS had not been reloaded as of yet, just waiting to see of all of the changes on this machine resolve the issue or not.
Thanks for your help...

 

[SlasherIT]

Good job on fixing it. Kudos to you for taking the initiative to take the DIY approach.

 

[Cr00zng]

Thanks Slasher...

Well, it was the DIY or the SEB (Send Everything Back) approach. Since I did need the machine, there wasn't much of a choice.

<runt>
One thing I've learned during this process that drivers and software are bloated piece of junk, even Intel's. While I can understand that they are trying to make the software more useful by adding more feature; sure, you go ahead and make it better. What is inexcusable that if the new feature doesn't need to run, they keep running it and lockup the system instead of exiting gracefully. That just bad programming that causes anxiety to people and they end up spending more than they'd need to.

Microsoft should've locked down the Windows kernel even more to prevent crapware causing these sort of issues. I just wish they'd have a video driver to replace the crappy ATI/AMD driver; might as well they are at it, please add a driver for the Marvell chip as well. I'd be ecstatic if my machine would use MS drivers only.
</runt>

Well, then it would be a MAC, wouldn't it

 

[SlasherIT]

Well, ATI graphics cards have a reputation for having really crappy drivers. This is one of the many reasons I am gonna stay away from ATI Graphics cards. Although AMD CPU's are good. Yes, that is something bad about Windows computers. Everything installed will cause bloat. Example, NVIDIA Driver's. When installing NVIDIA drivers, it come's with bloat such as NVIDIA 3D Stereoscopic, even though I don't need it because I can't use it. It installs with it a useless process that runs automatically, which consumes 10-20mb of RAM consistently. This service is only used to run NVIDIA Control Panel, without it, it won't run. What a useless process. Yes, drivers are crappy and annoying. Period.

Well, good luck on your computer, hopefully it will run well again.

BTW, since your problem is solved, it would be wise to lock this thread.

 

[Dave76]

If you're still having the random freeze issues, usually when you have more than 4GB RAM installed you need a integrated memory controller voltage bump.
I believe yours is around 1.0v, try bumping it to 1.15v.
The same i-series rule applies, don't go past 1.2v.

This might help, can't hurt.

 

[Cr00zng]

Well, I am not sure that there's is a setting in the Asus P8P67 LE board that would allow setting the voltage for the integrated memory controller. While there are options in the BIOS to set voltages for different components, none of them has values as you described.

The CPU has 1.8V auto adjusted and cannot be set in the BIOS directly by entering different values; however, it can be "offset" via the "+" and "-" keys in 0.005 increments.

The VCCIO, and PCH voltages can be adjusted from 0.7350 to 1.6850V in 0.005 increments. The DRAM voltage is for the memory slots and the voltage ranges between 1.185 and 2.1350V with the warning of not to exceed 1.65V.

That's about all of the voltages that can be modified.

The current settings are the default values in the BIOS via pressing "F5"; based on Asus' take on it, those are the optimal settings for the system. I have no intent to OC this system; it's plenty fast as it is.

The last couple of days there had been no issues whatsoever. While the memory had been changed from Crucial to Gskill, removing the ICH10R drivers and related Intel programs did actually resolved this issue. The chances are that the Crucial memory would work just as well with the ICH10R removed. Oh well, at least the WEI for the memory increased from 7.6 to 7.9.

 

[Cr00zng]

Um, this might be a stupid question, but...

How do you lock the thread?