• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

System fails randomly when doing something since CPU swap

B

bmwracer67

New Member
Joined
Sep 29, 2010
Messages
4 (0.00/day)
Hi.

I have a problem with my computer. I recently swapped my computer's old CPU for a new one. The old one was AMD Phenom X3 8750, the new one AMD Phenom II X4 965 BE. After the swap, the computer booted up and everything worked. Just the temps were a bit high for the cpu(40c idle and under load would go over 65c) so i needed a solution for the high temps. I undervoltaged the CPU a bit(from 1.35V to 1.275V) and ran stress test. Everything seemed fine, no errors and temps didn't exceed 62c. Soon after that there was a message that some process is not responding any more and needs to close. Very soon after that the computer just randomly rebooted. Next day it also rebooted this way without any warning or blue screen once or twice. I remember at these times the CPU had some load(a game running that utilized 1 core).

So just in case it had been from the CPU voltage, i upped it to 1.325V. I haven't had that random reboot since i did that(yesterday), but today when playing Live For Speed the screen just randomly went bad. I mean, the screen was just stripes in strange colours.

The system seems stable when idle, it appears, and may not fail for a long time when playing something.

Here are my system specs:
AMD Phenom II X4 965 BE @3.42GHz(stock clock rate, no OCing done)
ATI Radeon HD 4850 Passive(with a small fan beneath it for better cooling)
3.25GB of RAM(4MB in reality), 800MHz
Windows XP SP3
Motherboard Gigabyte GA-MA770UD3 rev.2.0

Here are the BlueScreenView reports from my 2 random reboots:
==================================================
Dump File : Mini092810-01.dmp
Crash Time : 9/28/2010 4:32:36 PM
Bug Check String : NO_MORE_IRP_STACK_LOCATIONS
Bug Check Code : 0x00000035
Parameter 1 : 0xba3a84cc
Parameter 2 : 0x00000000
Parameter 3 : 0x00000000
Parameter 4 : 0x00000000
Caused By Driver : WmFilter.sys
Caused By Address : WmFilter.sys+4cc
File Description : Logitech WingMan Hid Filter Driver
Product Name : Logitech Gaming Software
Company : Logitech Inc.
File Version : 5.09.129
Processor : 32-bit
Computer Name :
Full Path : C:\WINDOWS\Minidump\Mini092810-01.dmp
Processors Count : 4
Major Version : 15
Minor Version : 2600
==================================================


==================================================
Dump File : Mini092810-02.dmp
Crash Time : 9/28/2010 10:05:15 PM
Bug Check String : PAGE_FAULT_IN_NONPAGED_AREA
Bug Check Code : 0x10000050
Parameter 1 : 0xffffffff
Parameter 2 : 0x00000008
Parameter 3 : 0xffffffff
Parameter 4 : 0x00000000
Caused By Driver : Ntfs.sys
Caused By Address : Ntfs.sys+3f45d
File Description : NT File System Driver
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 5.1.2600.5512 (xpsp.080413-2111)
Processor : 32-bit
Computer Name :
Full Path : C:\WINDOWS\Minidump\Mini092810-02.dmp
Processors Count : 4
Major Version : 15
Minor Version : 2600
==================================================

Any help would be appreciated
 
Check whether there is an update for the bios of your motherboard.

Edit: Oh and welcome to TPU! :toast:
 
I wonder if the 2 blue screens are even related?

Here is info on the IRP Stack error ...
This problem occurs because the Mup.sys driver assumes that there must be no more than three file-system filter drivers running at the same time. The Mup.sys driver handles Distributed File System (DFS) file I/O requests. If there are four or more file system filter drivers, the I/O request packet (IRP) location buffer that is pre-allocated by Mup.sys will overflow. When this occurs, you receive the Stop error that is described in the Symptoms section.
Click to expand...

There is a hotfix available from here

page_fault_in_nonpaged_area info
 
Stock voltage for C2 965BE is 1.425v.

Stock voltage for C3 965BE is 1.4v.

If this cpu was bought very recently, there's a good chance you will need more cpu volts. 965BE is not longer the best AMD cpu(970BE is launched, 975BE inbound), so don't expect too much.

But definately give the other ideas here a go. I will say, however, that under the stock cooler, I cannot get 3.6ghz out of my 965BE...it gets too hot. Under TRUE cooling, max is 3.65ghz, beofre it gets too hot.

I am also curious if there is a little gold "N" or "T" on the cpu pcb...I bet it's an "N".
 
thanks for the replies. This seems to be a great forum.

I currently have the latest version of BIOS so that shouldn't be the problem.

About the voltages, mine had 1.35V by default. And yes it has been bought very recently, under a week ago. I have read that the 965 BE can be stable at even 1.24V with stock clocks. But then, different specimens differ.

My CPU temps are too high. I have this cooler: http://zalman.com/ENG/product/Product_Read.asp?idx=277

I did some CPU and GPU stress testing and both work well on 100%(or near 100%) load. I will put the voltage back up to 1.35V and then see what happens.

Thanks again


Edit: increased the voltage to normal. Now my temps are super-high, a few seconds of load with case closed will result in temps over 65c which i don't want to let them be. I wonder if it's my cooler or did i put too little thermal paste... But i think it aint the temperatures that caused the fails, at these times the temperature was always 50-55c

EDIT: My system failed again and I got the blue screen saying the cause of it was that ati2evxx which ended up in an infinte loop while i was playing Cities XL. :S
 
Last edited:
sorry for double post but I really need help. This does not make sense to me but maybe it does to you. I ran CPU stress test - 100% pass, memory is also good and video card stress test also went well. But when I play games, the system may just randomly crash. It never has crashed when idle.

I have got 4 BlueScreenView reports(2 of them are up in the first post) up to date and though it says that every time the error was caused by a different driver i see that in every bsod case one of the red-coloured filenames is ntoskrnl.exe. And once it was there on its own. So this is a common link between all the bsods.

Could it be the cause and how could I fix it? :S
 
bmwracer67 said:
sorry for double post but I really need help. This does not make sense to me but maybe it does to you. I ran CPU stress test - 100% pass, memory is also good and video card stress test also went well. But when I play games, the system may just randomly crash. It never has crashed when idle.

I have got 4 BlueScreenView reports(2 of them are up in the first post) up to date and though it says that every time the error was caused by a different driver i see that in every bsod case one of the red-coloured filenames is ntoskrnl.exe. And once it was there on its own. So this is a common link between all the bsods.

Could it be the cause and how could I fix it? :S
Click to expand...

There's actually a wiki page dedicated to that .exe

ntoskrnl.exe (and ntkrnlpa.exe on systems with Physical Address Extension support) is the kernel image for the family of Microsoft Windows NT operating systems. It provides the Kernel and Executive layers of the Windows NT kernel space, and is responsible for various system services such as hardware virtualization, process and memory management, etc., thus making it a fundamental part of the system. It contains the Cache Manager, the Executive, the Kernel, the Security Reference Monitor, the Memory Manager, and the Scheduler, among other things
Click to expand...

Have you tried reinstalling your OS?, also see if resetting CMOS will do anything.
 
what size is your powersupply?
just curious;)
 
CPU contains cpu core, cache, and memory controller. testing all three can be hard.

Personally, i'd look at PSU, plus cpu/mem.

Sounds to me like too low cpu volts.
 
Low cpu volts sounds about right. Although it could be temps. What were the temps on the CPU whenever it crashed? I remember you saying 62+c one of the times, which if I remember right is the upper limit for the quads from AMD, the HEY DO NOT PASS THIS temp. at least on my 955 it is. So if that is the case an aftermarket cooler should fix you up. If you have the space, the Xigmatek Dark Knight is a great cooler for AMD Quads, I have my 955 clocked up to 3.63ghz(Im a newbie to ocing), and it won't move an inch over 50c under prime95 for around half an hour, and dips down to 32c idle. Great cooler, but quite expensive, you get what you paid for though.


Do correct me if Im wrong in this post.
 
did you update BIOS??

i didnt get time to read the whole post.
 
Yes, he did.

In my opinion, if everything works well tested separately and crashes when used all at once, that might mean that your PSU might not be able to power all parts simultaneously.
 
^ that was my secoond opinion. you took it away :cry:
 
I'm very sorry, but it has appeared that memtest does make the system fail. I just did it(i thought i had done it with this cpu but i hadn't).

After about 5 seconds of memtest the computer just reboots. If floppy is not taken out of the drive then it just keeps booting to memtest and runs it for 5 seconds with no errors shown up to date and just reboots. When I go to bios just after that i see cpu temperature of about 43c. I upped the cpu voltage by 0.025V to test if it would increase the time that memtest could run, but it had no effect. Increasing memory voltage by 0.02V also didn't affect the results in any way. Also disabling USB legacy support didn't alter the results.

And about the temps - yes they are too high. I think I have to install my cooler again and put more thermal paste as last time i might have put too little. It's an aftermarket cooler.

And my PSU is Corsair HX650W

Edit: THE PROBLEM IS SOLVED :)

Eventually it came out that one of my two DDR2 modules was faulty. I took it out and now everything works fine :) I will need new RAM though
 
Last edited:
You must log in or register to reply here.
Back
Top