Hello

Recently I have gotten sudden BSODs while gaming the latest titles, such as
Batman Arkham Asylum (attached a minidump file)
Resident Evil 5
Ghostbusters
It happens after about 5 minutes of gaming (batman) or even several seconds (ghostbusters).

While I can play for hours normally titles like
Red Faction Guerilla
UT2004 and some older games

I am using Windows XP SP3 with the latest windows update hotfixes, latest sound and graphic card drivers (190.62).
Corsair 620HX PSU
Q6600
8800GTS 512
4 sticks x 1 GB OCZ RAM both dual channel (XP recognizes about 3gb)
3 Hard Drives
ASUS P5E MB with initial BIOS (I think it is '0107' Bios, date:12.10.07)
no overclock

I have this system for 2 years and did not have similar BSODs while gaming, in the past.

I will post any more information you require to help regarding this situation.

-----Things I tried:
chkdsk for all HDs
Can't be related to CPU or GPU overheating
Will add if necessary

Thank you.

Usually when we see game crashes occuring and at no other time it is a strong indicator of video card drivers.

But this one is different. Your error code is 0xA and these are caused either by hardware or drivers attempting to address a higher IRQ Level than they are designed for.

In your case the minidump specifically cited memory corruption as the probable cause. I suggest the following:

1. At MajorGeeks.com they have a free Video Memory Stress Test. Run it.

2. Run Memtest on your RAM to determine if there is indeed corruption.

See the link below and follow the instructions. There is a newer version than is listed but either one should work. If you need to see what the Memtest screen looks like go to reply #21. The third screen is the Memtest screen.

Let it run for a LONG time. The rule is a minimum of 7 Passes; the more Passes after 7 so much the better. There are 8 individual tests per Pass. Once all 8 tests are done then one Pass has been completed. Another 8 Tests will statrt again and though they look like it just repeating the same thing it is not.Many people will start this test before going to bed and check it the next day.

If you have errors you have corrupted memory and it needs to be replaced.

Also, with errors you need to run this test per stick of RAM. Take out one and run the test. Then take that one out and put the other in and run the test. If you start getting errors before 7 Passes you know that stick is corrupted and you don’t need to run the test any further on that stick.

Link: http://www.techspot.com/vb/topic62524.html


* Get back to us with the results.

Thanks for the reply

So far I did the Video Memory Stress Test, with several options. Check the log the results and for my comments on which options I had on or off (Ignore colors bit mask, Use onscreen buffer)

In the next days I will let Memtest86+ V4.00 (22/09/2009) run to test each stick of memory individually. And will post back with the results.
Is there any option I need to turn off in BIOS regarding memory in order for the system and memtest with only 1 stick of memory to work normally? Or the default options are fine?
EDIT: I understand that I need to run memtest per only 1 stick of RAM until testing all 4, and not run the test with 2 or more sticks as well, yes?
And does memtest create a log or do I just have to confirm that there are no errors after an overnight test by looking at the screen?

I remember that the memory (OCZ2P800R22GK) hsa a timing of CL 4-4-4-15, and I might have them set to 5-5-5-15 instead (default BIOS values).

You can run this test with all four sticks installed and even if you only have one at a time installed you don't have to worry about the BIOS. The screen will give the information you want and the two that are important here are Errors (if any) and number of Passes (again the minimum is 7 but beyond that so much the better).

For diagnostic sake set the CAS timings to OCZ specs and see what happens. You can always set it back to the current timinings if you need to.

Here is a result of one of the four sticks. If the options are ok I will do the same with the rest.

I changed the timing back to 5-5-5-15 (I had them manually set to 4-4-4-15) but I will keep them at 5-5-5-15 after the test. I assume if a memory stick has errors than this small thing will not matter.

Also after saving changes in the bios, the computer will turn off for 2 seconds, then turn back on. The same thing happens when simply turning on the computer from an off state. It was always like this I assume this is normal

Here are the results of each of the four memory sticks. (3 images here, 1 attached to my earlier post)
As you can see, there are no errors reported.
But I did notice something strange: see from 2 of the 4 pictures (each picture for one of the four memory sticks) are written DDR3 under Chipset.
All of the memory sticks are DDR2 of the same type and brand. Also the P5E MB has 4 memory slots all of DDR2, none of DDR3. Obviously I could not put a DDR2 stick in a slot dedicated for DDR3.
Note that every time I removed and put a new memory stick to test, I did put it in a different slot, eventually using all 4 slots.

So there are the results of each of the four memory sticks, also the Video Memory Stress Test log in an earlier post (does it report any errors or problems?).

What is next? Currently im using only 1GB of memory and I think to start a game and check if I get a BSOD, just to confirm it is indeed not memory related. But what do you suggest next?

Thanks

Now that is strange and yet at the bottom it reads it as DDR2. By the way, kudos for you for the amount of Passes you ran your memory. Too many people don't even have patience to run the minimum required 7.

Let's take this from a different angle. You do have the latest video card driver. Have you tried rolling back to an older vid driver? And, are still getting BSODs with the particular games you mentioned?

What is ECC? Is it ok that its reported as off in memtest screen? Are the FSB and RAM frequency OK? I know the RAM is 800MHz but im running it at 667MHz. Could any such settings be the cause?

Could the reporting of DDR3 in memtest screen be some bug?
I have a P5E MB with 4 DDR2 slots. A MB called P5E3 has 4 DDR3 slots.
I could put one stick of memory into each of the 4 slots to see on which of the slots, memtest reports DDR3 under Chipset info. If that is needed.

About installing RAM into the MB memory slots, I have to install the memory while the computer case is standing, and even when I apply moderated force, the motherboard is being forced back due to the force, it just worries me I will harm something.
I feel it takes too much force to push a memory stick into a slot, or maybe its much easier when the case is laying and not standing?

I will do more testing later with 1GB of memory and with current and then with older drivers.
I hope games that require more memory will not crash Because I have only 1GB right now?

I looked up your motherboard and it is definitly designed for DDR2 memory. Even the one that strangly reported DDR3, as I said earlier, it still reports at the bottom it is DDR2.

Dont't concern yourself that ECC is turned off. It has nothing to do with your type of RAM.

As for putting RAM into the slots it needs to be put firmly in place but NOT forced. The clips at both ends should snap closed if done correctly.

Keep in mind XP is only a 32-bit OS and will recognize only 3.5 gigs of RAM. To get the full use of 4 gigs you need a 64-bit OS.

* Do you know how to access your BIOS? If so see if your RAM is set to your manufacture's voltage specs.

I just ran batman and played for a while and no BSOD.
The game sure had many slowdowns I guess cause of the lack of memory, I use now 1GB on the most right, black memory slot.
I did also play Resident Evil 5 for a while and it runs fine and no BSOD. I forgot to add it to the list in my first post.
I think I can safely say that with the current hardware and no changes to software, games run fine.

I did not change drivers or anything aside from using only 1GB of memory.
Could one of the memory sticks still be flawed, or maybe a memory slot?

I guess the next step is to add another 1GB of memory, until finding the one that will cause a blue screen?

I just added another stick of memory so now I have 2GB installed, and will see if games run ok or not.
If I do get a BSOD like when I had all 4 sticks, I will remove the newely installed memory, and put another one instead.
Eventually finding which stick/s (or maybe slot/s) is causing the BSOD.
It is safe to assume it is memory related right?

As for installing the memory, I do have to apply a bit of force, or push it with my thumbs, in order for the memory to get into the slot. That force is also pushing the motherboard (and cpu heatsink) inwards a bit (or so it seems to me). I just wanted to know if its normal and can't damage. The case is standing as I install memory.

The recommended voltage for my memory is 1.9-2.1v. I do not know how much is set in the BIOS. I think that option is set to auto. Please tell me which option it is, and if I should set it to manual. I included some photos of the BIOS and of an application HWMonitor.
Note that I set 'DRAM Timing Control' from Auto to Manual, just to manually set, and make sure the CL is set to 5-5-5-15.
Also is there an application which can show me the memory voltage?

Update: After installing the 2nd memory stick and 2 hour of playing, I got a BSOD. Windows did not generate a minidump file! Could it be because I have 5 older minidumps in the Minidump folder?

Check the DRAM voltage. Don't worry about anything else. What does it say your memory is set at?

Also, no, you can have many more minidumps than 5.

As for installing memory, if you are careful you can lean the case to the side. I even installed it while laying on the side. There will be some "give" with the mobo. The trick is not to force the memory.

One other thing, Asus makes great boards but they are quite well known for being very picky about the RAM installed and the correct configuration.

Question: Is your RAM recommended for your motherboard according to Asus? Go to their website or a do a search for your particular model and see.

ok I checked the DRAM Voltage line in the BIOS.
It is set to auto. I am not sure what it means (how much voltage auto means).
Looking in the MB manual, it only says that the values range from 1.80v to 2.66v with 0.02 interval.
I Can set numbers there, but currently its set to auto as it was always.

I now removed the stick of memory that I had a BSOD with, and put a different one.
Shall I continue the test to check if I get another BSOD with more memory sticks? Or you think I should change some setting?
And could there be a memory slot that is damaged, or can it only be a memory (RAM) stick?

Edit: After I put a different stick in the same slot, I got another BSOD after 30 minutes.
Again there was no minidump created, why?
I can disable automatic restart and write the error when it happens if you want?
So now either 2 memory sticks are bad, or the memory slot is bad, or something else? whats next?

Edit 2: still with 2GB, with the stick that gave me a BSOD in earlier paragraph, 2 hour of gaming and now no BSOD

Edit 3: I have my pagefile in a different HD than the one windows is installed on.
Could the lack of pagefile in the HD windows is installed on be the reason there was no minidump created?
I set the pagefile on an HD with just data on purpose. Should I change it to be on the windows HD? or should I just create a 2nd small pagefile on C in order for minidumps to be created?

Edit 4: I kept my main pagefile but created a 128mb pagefile in C as well just for minidumps to be created, Is that OK?

And I think that the memory did not cause the last BSODs, so I will keep checking with a third memory stick (unless you tell me otherwise regarding voltage or what not)
I think the few latest BSODs were caused by the GPU, because I saw in the BSOD screen 'nvdisp.dll'
I attached that minidump, and weird that the BSOD happened while playing and when I had FRAPS on, I clicked the hotkey to take a screenshot and at the exact second the BSOD happened.
Please look in the attached minidump and see if its Different than the one I attached in the first post, and related to the GPU?
If so than I might also have problems with the GPU?
The -02.dmp was also caused while pressing the hotkey to take a screenshot, is it also related to the GPU?

As for memory recommended with my MB, I found the list in asus support page, and OCZ is not listed at all. I did use this computer for almost 2 years now though.
I don't think that if a memory model is not listed there, that it means it will not work with the MB. The list not so long.

To make it short, I currently have 3GB installed and games run fine.

I have 1GB out and I suspect it is the problematic but I can try to replace it with an installed one, and confirm a BSOD.
Does that sound right? Even though all of the memory sticks passed memtest.

I do not know why Batman crashed with FRAPS on but I hope its a software thing, I will install newer
versions of some of my programs and see if it will happen again.

I do wonder about my pagefile question if you have an idea.
Also should I manually set the DRAM voltage or leave it set to auto? I hope its not getting less than recommended while on auto.

Leave your voltage on auto most Bios will set it up for you !. & as for a crash whilst runnig fraps i would worry about it !.