Sign in with
Sign up | Sign in
Your question
Solved

BSOD happens daily

Last response: in Windows 7
Share
July 23, 2012 3:10:12 AM

Can some one please tell me what is causing these events and hot to fix them?

Image:

m
0
l
Related resources
a b $ Windows 7
July 23, 2012 4:58:34 AM

Pagefault in a nonpaged area is usually a memory error. Have you run memtest? Hardware?
m
0
l
July 23, 2012 5:18:59 AM

i ran the windows memory diagnostic and it came back with no issues
m
0
l
a b $ Windows 7
July 23, 2012 5:29:48 AM

Memtest runs outside of Windows and is better. 8 to 10 hours for 8GB. Any errors at all means something is wrong. Hardware?
m
0
l
July 23, 2012 5:43:17 AM

There are many "memtest" when i google it. can you please provide a link to the one i should use
m
0
l
a b $ Windows 7
July 23, 2012 5:47:19 AM

http://www.memtest.org/

Download and burn the .iso to a CD. Change the boot order in BIOS so the DVD/CD drive is first and boot off the CD. This allows all the RAM to be checked with none reserved for Windows. If you need a burning program ImgBurn is a great free program.

http://www.imgburn.com/
m
0
l
July 23, 2012 5:49:59 AM

ok thanks. i will post back with the results
m
0
l
a b $ Windows 7
July 23, 2012 5:50:37 AM

Wow 32GB? Is your motherboard even compatible with 32GB kits? That's why I keep asking what your hardware is so I can check things like that. New build? Has it always crashed?

What do you need with 32GB? You running a RAM disk?

You may need a voltage bump to run that much memory on any platform.
m
0
l
a b $ Windows 7
July 23, 2012 5:52:15 AM

Oh and .... 8 to 10 hours of memtest is what's recommended for 8GB. 32GB may need more like 20-24 hours.
m
0
l
a b $ Windows 7
July 23, 2012 1:21:22 PM

I'm not likeing the fact the he's getting 0x101: CLOCK_WATCHDOG_TIMEOUT, since that basically means the CPU is getting locked up...

That being said, the 0x50:p AGE_FAULT_IN_NONPAGED_AREA BSOD indicates RAM as the primary culprit...

Are you SURE the motherboard supports 32GB? Even if it does, you might need a voltage boost on the RAM/NB in order to gain stability.
m
0
l
July 23, 2012 5:21:58 PM

ok so i started memtest last night before i went to bed and woke up this morning to it still running. it currently has 169472 errors and has been running for 11 hours.... how much longer should i allow it to run. as well will it save the results some where after?

also here are my machine specs. sorry for not postiong them earlier.
Graphics Card: Asus HD7970 GDDR5-DC2-3GD5

Memory: Gskill Ripjaw x 32gb (4x8gb)

PSU: CoolerMaster Silent Pro Gold series 1000w ATX

CPU: AMD FX-8150 AM3+ 3.6Ghz

MOBO: Gigabyte GA-970A

OS: Windows 7 Professional 64 bit
m
0
l
a b $ Windows 7
July 23, 2012 6:20:44 PM

Even 1 error means something is wrong with your RAM.

You might try increasing the RAM voltage to 1.55v or 1.6v. 32GB also might require a BIOS update.

You may also have a bad stick but since it's 32GB I really think the problem is with the amount. Many boards have problems with such a large amount and not to mention the issues with 8GB sticks being compatible.

Are you actually using 32GB for something specific?

If a voltage bump does not help pull all but one stick and run memtest on each stick individually. If only 1 stick is producing errors then you have a bad stick.
m
0
l
July 23, 2012 6:25:21 PM

anort3 said:
Even 1 error means something is wrong with your RAM.

You might try increasing the RAM voltage to 1.55v or 1.6v. 32GB also might require a BIOS update.

You may also have a bad stick but since it's 32GB I really think the problem is with the amount. Many boards have problems with such a large amount and not to mention the issues with 8GB sticks being compatible.

Are you actually using 32GB for something specific?

If a voltage bump does not help pull all but one stick and run memtest on each stick individually. If only 1 stick is producing errors then you have a bad stick.


ok i will try increasing the volts. my bios has the most recent update. and no im not using it for anything specific, just figured it would help out in the future. i will rum memtest to each stick. if 1 of them comes back with errors what do you recommend i do? RMA?
m
0
l
a b $ Windows 7
July 23, 2012 6:33:33 PM

Yep. If a single stick causes the errors it's bad. If not then you still may have to RMA as that motherboard may not like 32GB.

Even 16GB is pretty much overkill unless you use it as a RAM disk or work with huge image or video files all day. For a gaming computer 8GB is fine. In fact only a few games do better with more than 4GB.
m
0
l
July 23, 2012 6:34:29 PM

so i upped the volts to 1.055 and pressed f10 to save then it restarted and nothing.... its like it is dead. when i push the power button the fans will start up but nothing is displayed on screen. no bios flash screen or anything just black....
m
0
l
a b $ Windows 7
July 23, 2012 6:37:11 PM

1.055 or 1.55? 1.055 is not enough voltage to boot. Clear CMOS and it should fix it.
m
0
l
July 23, 2012 6:44:29 PM

uh oh i must have done 1.055! how can i clear the cmos when i cant even get into the bios. when i push the power button its just a black screen.
m
0
l
a b $ Windows 7
July 23, 2012 6:51:44 PM

You clear CMOS manually on the board with a jumper. Check your manual for where it is or just use a flashlight. You can also remove the battery for a few minutes and that should do it too.

m
0
l
July 23, 2012 7:12:43 PM

ok great its up and running again. THANK YOU!! so i also upped the volts correctly to 1.55. what do you recommend i do now that they are on a higher voltage?
m
0
l
a b $ Windows 7
July 23, 2012 7:18:27 PM

Try memtest again. No need to let it run any longer than 1 error. If 1.55v does not do it try 1.6v. If that does not do it try running memtest on each stick individually.
m
0
l
July 23, 2012 7:20:30 PM

ok thanks. will post back after all that is completed
m
0
l
a b $ Windows 7
July 23, 2012 7:29:47 PM

You may also need to boost the NB voltage. Now here is where my not overlocking any AMD chips for the last 7 or 8 years comes into play and I have to say I can't really help you with that. Hopefully someone with an AMD system will chime in here and tell you what is safe.

You could also start a new thread titled something like " 32GB on a Bulldozer platform RAM errors." Something like that would get people with more AMD expertise than me to look at it. Just include full system specs and either a link to this thread or better a summary of what is in this thread and a link. The more detail you give people the more they can help generally.
m
0
l
July 23, 2012 7:53:16 PM

allright i will do that. when i ran the test on 1.55v as soon as it hit test 5 about 3000 errors showed up. then on 1.6v on test 5 as well only 8 showed up... any thoughts?
m
0
l
a b $ Windows 7
July 23, 2012 8:51:16 PM

Sounds like the right track. Maybe try a NB voltage bump as well. You can find the safe range with google. As I said I'm just not familiar enough with AMD chips to tell you there.
m
0
l
July 23, 2012 11:28:22 PM

Thanks for all your help. So do you think the problem is the voltages? i shouldnt have to rma the ram?
m
0
l
a b $ Windows 7
July 24, 2012 12:14:37 AM

Just based on how hard it can be to get 32GB to run correctly my guess is that it's the size of the RAM rather than the RAM being bad.

We have a member here on the forums much smarter about memory related issues than I am. His screen name is jaquith. My advice is to send him a PM and link him to this thread. If anyone on here can get you going it would be him.

Here is his user profile with a link to send private messages.

http://www.tomshardware.com/forum/profile-388864.htm

m
0
l
July 24, 2012 12:22:25 AM

Again thanks for all your help!
m
0
l
a b $ Windows 7
July 24, 2012 12:27:20 AM

No problem at all. I see people are trying in your new thread but so far no one I would call a really experienced builder has replied yet. Lol I have been building computers longer than they have been alive!

Try jaquith.
m
0
l
July 24, 2012 12:30:39 AM

hahaha i sent him a message, lets hope he replies.
m
0
l

Best solution

a b $ Windows 7
July 24, 2012 1:28:14 AM

Memory: Gskill Ripjaw x 32gb (4x8gb) F3-1600C9Q-32GXM DDR3-1600 9-9-9-25-2N @ 1.50v
ref - http://gskill.com/products.php?index=494
CPU-z - http://h9.abload.de/img/memoran5.jpg
CPU: AMD FX-8150 AM3+ 3.6Ghz
MOBO: Gigabyte GA-970A -UD3 or D3

--
BIOS Updates, there's been some noted stability issues so I recommend using the latest non-beta BIOS.
BIOS versions:
UD3 use F6
D3 use F9
--

First, I assume no CPU OC; a CPU can cause a problem and require different settings.

BIOS:
Load Optimized Defaults
Set Memory Clock -> Manual
Memory Clock -> 8 ; 8*200MHz = 1600MHz
DRAM Configuration /Enter
DDR3 Timing Items -> Manual
1T/2T Command Timing -> 2
CAS# latency -> 9 ; failure try 10 - if DRAM 1.55v + VID 1.25v fails
RAS to CAS R/W Delay -> 9 ; failure try 10 - if DRAM 1.55v + VID 1.25v fails
Row Precharge Time -> 9 ; failure try 10 - if DRAM 1.55v + VID 1.25v fails
Minimum RAS Active Time -> 25 ; failure try 30 - if DRAM 1.55v + VID 1.25v fails
-
DRAM Voltage Control -> 1.50v ~ 1.55v ; increase if errors
CPU NB VID Control -> 1.20v ~ 1.25v ; increase so that the actual voltage is 1.20v ~ 1.25v or +0.10v
Save & Exit = Yes

--

Updating the BIOS via Q-Flash:

Nice 3-part video:
1. http://www.youtube.com/watch?v=UCCId1Mcrw4
2. http://www.youtube.com/watch?v=Qlx5GBkiaM0
3. http://www.youtube.com/watch?v=EX78S4qyHu0

Nice informative link - http://forums.tweaktown.com/gigabyte/27576-bios-flashin...
Share
July 24, 2012 2:24:23 AM

Wow thanks a lot! i ran memtest and received 0 errors where before i received thousands. The only thing now is that my cpu tems are higher and it seems to be working harder, also log in time is slower and it doesnt seem as snappy as before. is that a result of what you had me do?
m
0
l
a b $ Windows 7
July 24, 2012 3:33:20 AM

Okay, Tons of errors = snappy vs No error = crappy...depends on your version of 'snappy.'

What exact changes did you need to do from the settings above?

The increase of 'CPU NB VID Control' is the culprit for temps, but it's what's keeping you stable; you can 'try' Auto but I'd test the RAM again.
m
0
l
July 24, 2012 3:41:13 AM

I changed everything. but ya at least its stable now. Thanks to both you and anork3, you both were a great help
m
0
l
July 24, 2012 3:41:58 AM

Best answer selected by marcus8h.
m
0
l
a b $ Windows 7
July 24, 2012 3:54:15 AM

If CPU NB VID Control -> Auto fails then 'try' +0.05v.

Good Luck! :sol: 
m
0
l
a b $ Windows 7
July 24, 2012 4:02:01 AM

Glad you got it running correctly. Told ya jaquith could help if anyone could!

Just for the hell of it I would still try and run memtest for at least 12 hours or so. Just to be sure.
m
0
l
July 24, 2012 4:05:32 AM

anort3 said:
Glad you got it running correctly. Told ya jaquith could help if anyone could!

Just for the hell of it I would still try and run memtest for at least 12 hours or so. Just to be sure.


Yea thanks for sticking with me throughout all this! i will run memtest over tonight. :) 
m
0
l
a b $ Windows 7
July 24, 2012 12:30:35 PM

Good to hear the improvement; had a suspicion it was RAM related...Once you get beyond 16GB or so, you get into that relm of motherboard incompatability or needing to increase voltage for stability.
m
0
l
!