Windows 7 Random BSOD's After CPU OC(?)

[Pekuda]

Hello!

Let's start with me saying that I am very thankfull to everyone who is willingly helping me with this problem!

To the point:

I get BSOD's after I do automatic overclocking to CPU and memory things, bluescreen has come on everything on factory setups so im not 100% sure the OC is the reason.

Here is few usefull link for my specs:
Normal:
http://speccy.piriform.com/results/umUwuC3MNTDz5jra5FEuaEB

Auto overclocked:
http://speccy.piriform.com/results/ifDfiVJO6P8Rx52683nvhNx

I've included minidump file in the attachments.

Thanks for all the help in advance!

 

[Mike]

Hi, I took a look at the error and it's a rather obscure one.

Date/time:              2/5/2012 3:14:40 AM GMT
Uptime:                 00:50:02
Machine:                
Bug check name:         [URL="http://www.google.com/search?q=MSDN+bugcheck+DRIVER_IRQL_NOT_LESS_OR_EQUAL"]DRIVER_IRQL_NOT_LESS_OR_EQUAL[/URL]
Bug check code:         0xD1
Bug check parm 1:       0x2
Bug check parm 2:       0xC
Bug check parm 3:       0x8
Bug check parm 4:       0x2
Probably caused by:     [URL="http://www.google.com/search?q=ntoskrnl.exe"]ntoskrnl.exe[/URL]
Driver description:     [URL="http://www.google.com/search?q=NT Kernel & System"]NT Kernel & System[/URL]
Driver product:         [URL="http://www.google.com/search?q=Microsoft® Windows® Operating System"]Microsoft® Windows® Operating System[/URL]
Driver company:         [URL="http://www.google.com/search?q=Microsoft Corporation"]Microsoft Corporation[/URL]
OS build:               Built by: 7600.16841.amd64fre.win7_gdr.110622-1503
Architecture:           x64 (64 bit)
CPU count:              6
Page size:              4096

Bug check description:
This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.

A commendable use of Speccy to supply pertinent information. Very thorough here.

Question: Does the system crash when it is NOT overclocked?

If the answer to that question is no, then quite simply, stop overclocking. You may be over-volting your CPU or RAM. Settings could cause serious system damage. It is the most common reason for this type of BSOD under the conditions you have described. If you do not know for sure, here is what I want you to do. Keep in mind, you may not like the answers because they are somewhat a pain, but you are an overclocker, so you are probably already familiar with the steps:​

Option 1: Use the PC extensively without overclocking it for a few days (72 hours minimal).

We need to determine if the system crashes when it is not being overclocked, or if the damage is a permanent result of you overclocking too much. To do this, we can do several things. You can use the computer, as you always do, for several days, optimally, regular daily usage for about 72 hours. You can come and go, go to work, do whatever it is you do, but run the system without any overclock settings and see if it ever crashes. If it never does, you have a good indication the system is stable.​

Option 2: Conduct burn-in testing when the computer is not overclocked.

To expedite this process, and put all chips (no pun intended) on the table, conduct burn-in testing when the system is not overclocked. What is burn in testing? When you perform a burn-in test, you operate the processor at 100% maximum load using a stress test / benchmarking software to ensure that the system does not crash. Burn-in testing is used after new systems are assembled to determine the reliability of the processor and memory. The initial testing is usually performed by manufacturers at companies like Dell and HP automatically using custom-built utilities during the assembly and configuration steps. IT consultants and ISPs that do professional hardware installs will often perform this step either on-site or off-site for the customer at their discretion (sometimes charging a premium fee). The way most system builders do it is through running an assortment of programs. This is a step consulting firms will often times charge s lot for with servers, primarily because it is easy to sell to businesses and they have no idea what it is other than that it sounds important. When this test is done on deployment to an on-site location, software is used, sometimes with additional utilities to watch system activity from the office, with something like RAdmin and PA Server Monitor. In your case, one of the most highly rated items for simply conducting the burn-in test for this task is prime95. This software stresses your hardware to its maximum capacity to determine if it can take the heat. Many laptops, even after being shipped as factory refurbished or brand new, will not pass prime95 if they are not placed on a cooling pad. This is due to the inability of a laptop to cool itself effectively after a prolonged period of time. It is the nature of miniaturized components; yet justifiable for a RMA in most cases. How quickly the laptop crashes is of particular concern. If the laptop crashes within a few minutes or hours of running prime95, this is an indication that there may be defects in the processor, the seating of the processor, the ability of the fans to adequately cool for even a short period of time, or other components like RAM. Branded and custom desktop computers should be able to operate prime95 indefinitely without crashing. Because this test stresses components, there is a risk of complete hardware failure when you perform this type of test. However, by passing complex instructions to your processor and stressing both the CPU and RAM, you will have a good idea if your system will blue screen a month from now by using prime95. I note that you already risk complete hardware failure by overclocking any way.

I recommend this method for so long as you know and accept the risks. Download prime95 and run the 'Torture Test' for at least 24 hours.​

If your system doesn't crash after 24 hours running prime95 chances are you are fine. If you can complete 48-72 hour run through, the system is likely to be completely stable. This is faster than just using the computer to see if it crashes and it is the preferred method of system builders and overclockers everywhere.

• Option 3: If you are unwilling to perform either Option 1 or 2, please remove all AVG products from your system using Control Panel -> Programs and then the AVG remover utility.
  

The reason that I suggest doing so is because 'avgldx64.sys' and 'avgidseh.sys' are a loaded device driver modules at the time of the BSOD. These files stand out because AVG is known to create problems within the hardware abstraction layer (HAL) on systems and can often times malfunction creating blue screens. When the entire Windows kernel is reported as the actively crashing component, this is usually the result of some piece of software that is not part of the Microsoft Windows operating system, such as an anti-virus utility, passing incorrect values a long to the operating system core. This is one of the few problems, outside of physical hardware defects, that will cause a computer to immediately halt. If you don't want to take my word for it, simply look up "AVG BSOD" on Google or any of the above files with "blue screen" after them in Google. You will find countless people having AVG crash their system. You may even want to uninstall AVG and go back to your overclocking settings and see if it still blue screens. I would recommend this, as you could always replace AVG with Microsoft Security Essentials, ESET Smart Security, or ESET NOD32. Like it or not, AVG has been known to cause blue screens in Windows 7 for some people. This often times depends on their hardware configuration and list of installed programs - in such a way that is not yet entirely understood or resolved by the developer to my knowledge.

​

Follow these steps, and please do us all a favor by letting the forums know if any of these solutions work for you. As you can probably tell, I have spent a great deal of time providing these solutions to you, and would like to know what the cause was. By exercising any of these options, you are on the right path. Rarely can a BSOD analysis point out the problem straight away, but when we hear about overclocking, that is usually the problem. When we see AVG in the loaded modules, that's another red flag. Some people will have constant BSODs, half a dozen or a dozen or more, and the problem is clearly identifiable by running memtest86+ or some other utility. Often times people do not want to come to grips with the fact that it might be a hardware problem. In your case I bet it is either overclocking or AVG. Now here are the tools to find out. Good luck!

 

[Elmer]

I'll go 100% with Mike, except I would have put the "remove AVG" first. Especially as you are running 64bit. Replace with BSOD friendly Link Removed due to 404 Error At least until your issues are resolved.