Prime95 Instant BSOD

[Meanbunny]

Hello. After reinstalling my operating system after having almost all the hardware in my computer replaced by dell, I was getting random shut downs every 15 minutes to 1 hour doing various tasks on the computer. I was told by a dell tech to try running a prime 95 stress test to see if the CPU might have something to do with it. When I run Prime95 Small FFT's test, I get an instant BSOD blaming ntoskrnl.exe+75c00.

Specs:
Processor: Intel (R) Core i7 980x Gulftown
Motherboard : Alienware x58 Chipset
Graphics: 2x ATI Radeon HD 5970 Crossfire
Hard Drive: 2x Western Digital 640gb Sata II Raid 0 Array
Memory: 3x 2gb 1867 Mhz Kingston Hyper X Sticks
Power Supply: 1000 Watt

 

[YoYo155]

Welcome, please follow the http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html .

 

[Meanbunny]

Thank you, I have removed the dropbox link and uploaded the requested file. Thank you greatly for your time and attention YoYo.

 

[PSCO2007]

Hello. After reinstalling my operating system after having almost all the hardware in my computer replaced by dell, I was getting random shut downs every 15 minutes to 1 hour doing various tasks on the computer. I was told by a dell tech to try running a prime 95 stress test to see if the CPU might have something to do with it. When I run Prime95 Small FFT's test, I get an instant BSOD blaming ntskrnl.exe+75x00.

Specs:
Processor: Intel (R) Core i7 980x Gulftown
Motherboard : Alienware x58 Chipset
Graphics: 2x ATI Radeon HD 5970 Crossfire
Hard Drive: 2x Western Digital 640gb Sata II Raid 0 Array
Memory: 3x 2gb 1867 Mhz Kingston Hyper X Sticks
Power Supply: 1000 Watt

This explains the error and possible solutions: NTOSKRNL.EXE is missing or corrupt.

Doesn't seem to apply to Win 7, but might point you in right direction?

 

[YoYo155]

Hi thanks of the files .
The bug checks varies a bit, but not one mentions Prime95 as the active as the active process.

Can you elaborate on which parts were replaced by Dell (and possibly if you know, for what reason)?
When a PC instantly reboots/BSOD's when starting Prime95 a few things come to mind.
Most common is high temperatures, the others are a PSU fault, an unstable overclock,
the least common are a faulty motherboard and / or CPU.
Are you sure that the PC bug checks and not just restarts as a safety mechanism, do you see the BSOD?
Are you monitoring the temperatures in any way?

• Download Speccy and post a Screenshot of the summary window.

As for the bugchecks, as I said they vary so will deal with the obvious first than continue
from there once the above (the reason that the PC can't handle Prime) is cleared.


You have a very old driver that doesn't do nothing but trouble regardless
of the other reasons the PC is unstable.

BugCheck 24, {1904fb, fffff8800a73f3f8, fffff8800a73ec60, fffff80002c80a7e}

*** WARNING: Unable to verify timestamp for SiWinAcc.sys
*** ERROR: Module load completed but symbols could not be loaded for SiWinAcc.sys
[COLOR=Red][B]Probably caused by : SiWinAcc.sys[/B][/COLOR] ( SiWinAcc+102e )

lmvm SiWinAcc
start             end                 module name
fffff880`01364000 fffff880`0136d000   SiWinAcc T (no symbols)           
    Loaded symbol image file: SiWinAcc.sys
    Image path: \SystemRoot\system32\DRIVERS\SiWinAcc.sys
    Image name: SiWinAcc.sys
    Timestamp:        [COLOR=Red][B]Fri Jun 15 03:02:02 2007[/B][/COLOR]

SATALink driver accelerator    c:\windows\system32\drivers\[COLOR=Red][B]siwinacc.sys[/B][/COLOR]

• Driver Reference Table - siwinacc.sys
  Silicon Image driver - comes with package for the SIxxxx drivers.

I found an updated version from 2010 but it's best you use your Service Tag to update the drivers.

• Dell Support

Post back with the results .

 

[Meanbunny]

*Please see the following post to help explain the Silicon driver problem*

Thank you for your time kind sir.

I will start by giving you a short summary of my recent PC problems. Originally my PC randomly stopped posting one day and so in an attempt to find out what was causing it, Dell ended up replacing everything except the hard drives (CPU, Memory, GPU's, Motherboard, CPU Liquid Cooling Module, PSU) until they found out it was a faulty CPU. After that the computer would post but I would get random shutdowns every 15 minutes to 1 hour doing random things like watching videos, playing games, installing programs from desktop, even when sitting idle it seemed sometimes. Since then I formatted and reinstalled windows and have just now recently updated all my drivers. I have not had a random restart in about 18 hours, but the tech wanted me to make sure that everything was working properly with the new CPU so he recommended I test with Prime95. He said I should be able to get through a Small FFT's test with Prime95 without any problems. When I start Prime95 Small FFT's test, I get an instant BSOD. Yes I can see the BSOD, it is not a debugging window. Its the same BSOD that come up in Blue Screen Viewer in the dump files. The main one is the ntskrnl.exe + 75c00. He said that if I can't get through a Small FFT's test, then something is wrong and it isn't build right, or its a possible software problem.

So pretty much I am trying to figure out why my computer can't run a stress test without BSOD'ing.

EDIT: The CPU is not overheating, the hottest I ever seen it get before crashing is around 65 to 70 degrees celsius when attempting to run the stress test. In fact the test doesn't even really start to be honest to even give it a chance to jump more than 10-15 degrees from its normal 38 to 40 degree celsius range. The monitoring software I use is HW Monitor and CPU-Z.

Here is the Speccy SS requested below.

 

[Meanbunny]

I have recently added 50+ more windows updates that I didn't know I was missing and windows is now fully up to date. After doing that I tested Prime95 and I am getting the same BSOD. I will upload an updated version of the SF folder requested on thread creation. Thanks again for your time YoYo. Hope to speak to you soon.

New news to report.

I just got my first random BSOD while just browsing the internet for new drivers after about 20 hours without one. What is even crazier is when I log back on, my minidump folder is EMPTY! What could have caused this? Luckily I was able to catch the debug code on the BSOD which seems to be the same exact code I get when Prime95 crashes. Its the 0x000003b one.

Okay, about the Silicon driver. The Silicon drivers that I downloaded from Dell don't seem to be installing right. When I click on the setup file it says windows configures 3132-W-D for about 2 seconds and then the window just disappears not giving any confirmation or message letting me know it installed right. I tried going to the Silicon website to update my drivers but the Only drivers I could find for the 3132 software I could find required a Bios and special card I think? I am having trouble finding the correct drivers, but when I look at device manager it says my drivers for it are up to date and it dates them at 06/24/2009. Under the properties under device location it says no device detected, and all the fields under device information are all blank. I do not remember having this controller on my system before. After installing what I thought were the wrong drivers I started getting a phantom SATA controller showing up under "Other Devices" and it kept failing to update the driver. So I did a system restore and I now for the first time am seeing a Silicon SATA link controller under my Storage Controllers for the first time saying that its working properly and the driver is up to date as of 2009. This is very confusing, I guess I am going to have to wait until you get back to me to figure out how to update the drivers for it. Also, is there even a way to confirm through my hardware that I even have this SATA Controller linked to my system? Since I don't have any PCI express card plugged in for any special raid controller, the only drivers I could find required a Bios and all that. I am going to go ahead and update my SF Diagnosis File to see if you see a change in the driver or if it was updated from 2007 to 2009 possibly. Sorry, just very confused at the moment. I probably need to get some sleep, heh.

For now, the biggest news is that I am now getting the random BSOD's again WITHOUT running Prime95 stress test just doing normal windows tasks. I suspect something wrong with the CPU they replaced the old one with. I think maybe the VID default core voltage was set too low at the factory or something because my BIOS is set to optimal defaults and I am not overclocked at all. Only other weird thing is my memory multiplier is set to 133x14 = 1867 Mhz. Why is it showing in speccy that they are running at 935 Mhz? I am assuming because of the Intel (R) Speed Step technology it is underclocking my RAM while the CPU is idle? I also noticed that even though my Vcore is suppose to be default about 1.12-1.2 V, it is sitting idle at about .936 V. Probably again a possible side effect of the Speed Step technology.

Some additional important information that you might find useful:

I have run chkdsk /f/r manually using the windows chkdsk checking both the repair and find bad files and sectors options.

I have also run Memtest86+ for 5 passes with each of my 3 memory sticks and received no errors. When I put all 3 sticks in and ran for 5 passes, I got 2 errors on test 7. I have heard that this doesn't neccescarily mean a memory problem, it could just be some bug that happens in certain situations? Not to mention it is brand new memory. I also doubt its the memory controller as it is also a brand new motherboard.

I don't know whats going on, but hopefully with your help, YoYo, we just might be able to get to the bottom of this. The last thing I wanted to do was to exchange my system with Dell if there was anything I could do about it. So if we can figure this out somehow, it would really mean a lot to me.

Once again, thank you for your time and effort YoYo. I hope to hear back from you soon.

Note: I also want to apologize as I have edited the last 2 posts extensively over the past couple hours as I have answered many of my questions I had written myself due to research and troubleshooting on my own. I apologize if it has caused any confusion if you had happened to read them during that time.

(Updated SF File below was from before the last random BSOD deleted my minidump folder)

 

[YoYo155]

Hi thanks for the added info .
From the last two bug checks I was able to verify that the Silicon RAID controller driver was not updated,
I wouldn't worry about it at the moment as there is something else I want you to perform as a test
in which I'll elaborate further later in this post.

At this point I suspect (even if they are new) the Motherboard, RAM, PSU, CPU and BIOS settings
as the possible culprits.

I think maybe the VID default core voltage was set too low at the factory or something because my BIOS is set to optimal defaults and I am not overclocked at all.

Please set the X.M.P profiles, where available, for the RAM mostly.
Some motherboards have a Safe setting as well as a Normal setting, if yours does
please set it to Normal.

Only other weird thing is my memory multiplier is set to 133x14 = 1867 Mhz. Why is it showing in speccy that they are running at 935 Mhz? I am assuming because of the Intel (R) Speed Step technology it is underclocking my RAM while the CPU is idle?

No, this is normal. That is how DDR or Double Data Rate works, 935*2= 1870 or ~1866.

I also noticed that even though my Vcore is suppose to be default about 1.12-1.2 V, it is sitting idle at about .936 V. Probably again a possible side effect of the Speed Step technology.

Yes that is normal as well, not so much a side effect.

I noticed that all of the bug checks (apart from one which was a TDR so it's not really a bug check)
occurred on processor #5.
Please try eliminating cores by starting the Prime95 test using only 1 thread and working your way up
until instability is reached.

View attachment 292178

If you still crash at the get go try eliminating the cores from Windows all together.
The disadvantage of this method is you'll have to restart to take effect each time a change is made.

rb: > Run > msconfig :
View attachment 292179

 

[Meanbunny]

Just got another random BSOD seemingly unrelated to the Prime95 ones. While I perform the tasks you have requested here is an updated SF Diagnostics File incase you want to have a look at it. Thank you SO much for your time YoYo. I think it has something to do with the CPU. Dell claims that CPU's cannot be refurbished and that it is new, even though it came in an Alienware box not an Intel box. Then again, possible the CPU is fine and there are something underlying causing it.

I will report back with my results. Thanks again YoYo.

 

[Meanbunny]

Here is my newest information.

I went into the Bios and apparently the XMP support is set to "AUTO". A funny thing I should mention is that when I enter Advanced DRAM Configuration, there was an option called Memory-Z to enter. When I opened it, there were 3 options, DIMM1, DIMM2 and DIMM3 Memory Information. DIMM1 and DIMM3 when I entered them showed all the info. When I tried to enter DIMM2, my whole Bios locked up and I could not ctrl+alt+delete out of it so I had to restart manually. To make sure it wasn't a fluke, I did it 2 more times with the same exact result.

Onto Prime95.

I was able to run up to 8 threads on the Small FFT's test, then Prime95 started to crash within seconds, although I was not getting any BSOD's. Each thread I added the crash got sooner and sooner until I finally got to the full 12 threads and I got the instant BSOD again.

After I set windows to run with only 1 core (no setting for no cores) and restart, what should I do next? Thanks again for all your help YoYo. Hope to hear back from you soon.

EDIT: I restarted with Windows using only 1 processor and Prime95 ran the same way it did only allowing 1 thread in the Small FFT's test. Even when I get to thread 8 when it starts to crash to desktop, I don't get any errors, just the application crashing. I have had the BSOD's happen with as little as 10 threads before.

 

[Meanbunny]

Just attempted to play Runescape because my sister wanted me to try to get on to at least do the 30 minute Halloween event and I can't even keep running long enough for that. Got another Bsod blaming Ntoskrnl.exe this time a different address. This time it is ntoskrnl+75bc0. Bugcheck code 0x0000003d.

 

[YoYo155]

Here is my newest information.

I went into the Bios and apparently the XMP support is set to "AUTO". A funny thing I should mention is that when I enter Advanced DRAM Configuration, there was an option called Memory-Z to enter. When I opened it, there were 3 options, DIMM1, DIMM2 and DIMM3 Memory Information. DIMM1 and DIMM3 when I entered them showed all the info. When I tried to enter DIMM2, my whole Bios locked up and I could not ctrl+alt+delete out of it so I had to restart manually. To make sure it wasn't a fluke, I did it 2 more times with the same exact result.

Good detective work! this is something we need to investigate further.
Start by taking out the corresponding module (stick) for that slot and stress with Prime95 without
that slot populated. Does it crash?

Onto Prime95.

I was able to run up to 8 threads on the Small FFT's test, then Prime95 started to crash within seconds, although I was not getting any BSOD's. Each thread I added the crash got sooner and sooner until I finally got to the full 12 threads and I got the instant BSOD again.

After I set windows to run with only 1 core (no setting for no cores) and restart, what should I do next? Thanks again for all your help YoYo. Hope to hear back from you soon.

EDIT: I restarted with Windows using only 1 processor and Prime95 ran the same way it did only allowing 1 thread in the Small FFT's test. Even when I get to thread 8 when it starts to crash to desktop, I don't get any errors, just the application crashing. I have had the BSOD's happen with as little as 10 threads before.

The way you describe it getting gradually worst sounds like it's a thermal issue still.
I tend to trust Speccy, but it won't hurt to get a second opinion.
Download HWinfox64 when you launch it it will ask you to choose what you want to load
select the Sensors only option.
Report the temperatures back .

 

[Meanbunny]

Sorry YoYo, been busy for the past couple days but am back to troubleshooting.

When I remove the DIMM2 stick I still get the exact same BSOD when running Prime95. I also noticed that when I try to run HWInfo after a few seconds of it being open, my whole system locks up and I have to manually shut down via power button. I am not sure if that happened when I ran it the first time before I removed the DIMM2 stick.

When I run Prime95 it crashes so fast there is no time to see any temperature change in the CPU. It all stays at around 50-55 degrees celsius. Even when I was running with less threads all the way up to around 10 threads where it was just crashing and not BSOD'ing the highest I got to was about 75 degrees celsius.

Also, I am getting BSOD's when doing light load things like watching youtube videos and playing internet browser Java games. The CPU is almost at idle temperatures when I am getting the BSOD.

I want to continue trying to address this problem if you don't mind, but in the mean time I want to tell you about the system dell offered me for an exchange for the one I have now and wanted your opinion if I am being ripped off or not.

Here are my current specs, followed by the specs of the one they are offering me, along with pricing.

Current Specs:

System: Alienware Area 51 ALX
Processor: Intel (R) Core i7 980x Gulftown @ 3.33 Ghz (2010 processor, Intel Website Price: $1000)
Memory: 3 x 2 GB 1867 Mhz Tri Channel Kingston Hyper X Ram Sticks
Graphics: 2 x ATI Radeon HD 5970 Crossfire
Motherboard: Alienware 0J560M Intel X58 Chipset
Hard Drives: 2 x 640 GB Western Digital SATAII Drives Raid 0 Array
Power Supply: 1000 Watts

New System:

System: Alienware Aurora
Processor: Intel (R) Core i7 4820K (Current Price: $350)
Memory: 4 x 4 GB 1600 Mhz Quad Channel Ram Sticks (Manufacturer Unknown)
Graphics: 1 x ATI Radeon HD 8990
Motherboard: x79 Chipset (Not sure of exact board)
Hard Drives: 2 x 3 TB Hard Drives Raid 0 Array
Power Supply: 875 Watts

When I got this system a family member had to pay $5000 for it and the system they are offering me now is worth around $2500. I am mostly concerned about the CPU and GPU. I just wanted to know also in your opinion if you think this sounds fair or if I am being ripped off. Thanks for your time YoYo. Hope to hear back from you soon.

 

[YoYo155]

I am out of ideas unfortunately, I'll see if someone else has any other thoughts .

EDIT:
By the way, did you clear the CMOS yet? all other overclocks are null right?

Do you have any use for rivatuner?
If you don't please remove it.

Reset the BIOS back to default:

• CLR CMOS

   Note

Write down the current value of the SATA Mode!
Its either AHCI or IDE.
After resetting the CMOS go back and verify the value for SATA Mode is what it was
when the OS was installed.

   Note

If any component is overclocked reset it back to stock speeds!

 

[Meanbunny]

Yes I have reset the CMOS a couple of times.

Some new info. Today I was tinkering with the Bios under the CPU settings and there were a couple settings called TDP and TDC override. They were set to 110 and 130 respectively. Just tinkering around, I set them to 120 and 140. I go back to Prime95 and amazingly no BSOD's. What in the world did I change that made it run Prime95 stable all of a sudden?

Beyond the whole Prime95 issue, what I don't understand is that I ran this computer for 2 years without a single BSOD before I got all my hardware replaced and now I can't even watch youtube videos or play Java browser games without BSOD'ing. I can understand Prime95 crashing because the CPU is not getting enough juice or something, but not while doing light load things like browsing videos and playing Java games.

If anyone else does have any input I would greatly appreciate it.

Also please let me know if you think this sytem they are offering me in exchange is a fair exchange or if I am being screwed over.

Current Specs:

System: Alienware Area 51 ALX
Processor: Intel (R) Core i7 980x Gulftown @ 3.33 Ghz (2010 processor, Intel Website Price: $1000)
Memory: 3 x 2 GB 1867 Mhz Tri Channel Kingston Hyper X Ram Sticks
Graphics: 2 x ATI Radeon HD 5970 Crossfire
Motherboard: Alienware 0J560M Intel X58 Chipset
Hard Drives: 2 x 640 GB Western Digital SATAII Drives Raid 0 Array
Power Supply: 1000 Watts

New System:

System: Alienware Aurora
Processor: Intel (R) Core i7 4820K (Current Price: $350)
Memory: 4 x 4 GB 1600 Mhz Quad Channel Ram Sticks (Manufacturer Unknown)
Graphics: 1 x ATI Radeon HD 8990 (OEM rebranded version of the 7990)
Motherboard: x79 Chipset (Not sure of exact board)
Hard Drives: 2 x 3 TB Hard Drives Raid 0 Array
Power Supply: 875 Watts

 

[YoYo155]

Yes I have reset the CMOS a couple of times.

Some new info. Today I was tinkering with the Bios under the CPU settings and there were a couple settings called TDP and TDC override. They were set to 110 and 130 respectively. Just tinkering around, I set them to 120 and 140. I go back to Prime95 and amazingly no BSOD's. What in the world did I change that made it run Prime95 stable all of a sudden?

Beyond the whole Prime95 issue, what I don't understand is that I ran this computer for 2 years without a single BSOD before I got all my hardware replaced and now I can't even watch youtube videos or play Java browser games without BSOD'ing. I can understand Prime95 crashing because the CPU is not getting enough juice or something, but not while doing light load things like browsing videos and playing Java games.

If anyone else does have any input I would greatly appreciate it.

Also please let me know if you think this sytem they are offering me in exchange is a fair exchange or if I am being screwed over.

Current Specs:

System: Alienware Area 51 ALX
Processor: Intel (R) Core i7 980x Gulftown @ 3.33 Ghz (2010 processor, Intel Website Price: $1000)
Memory: 3 x 2 GB 1867 Mhz Tri Channel Kingston Hyper X Ram Sticks
Graphics: 2 x ATI Radeon HD 5970 Crossfire
Motherboard: Alienware 0J560M Intel X58 Chipset
Hard Drives: 2 x 640 GB Western Digital SATAII Drives Raid 0 Array
Power Supply: 1000 Watts

New System:

System: Alienware Aurora
Processor: Intel (R) Core i7 4820K (Current Price: $350)
Memory: 4 x 4 GB 1600 Mhz Quad Channel Ram Sticks (Manufacturer Unknown)
Graphics: 1 x ATI Radeon HD 8990 (OEM rebranded version of the 7990)
Motherboard: x79 Chipset (Not sure of exact board)
Hard Drives: 2 x 3 TB Hard Drives Raid 0 Array
Power Supply: 875 Watts

Good news (?) .
My guess would be degradation over time perhaps :huh:.

About the specifications that Dell is offering you. the price point is not the only thing (although an important factor none the less) You can't really compare the two machines apples to apples sort of speak
the important thing in my mind is the so called 'real world performance' a good way to get a good
baseline is to research benchmarks and see how the (completely different generations and skews)
stack up against each other in your day to day usage case.For example an intel six-core, X79 chipset, CPU
would be awesome in productivity and will crush more mainstream CPUs in that respect but would suffer
where it comes to gaming.
It's up to you to make these considerations based on all of the mentioned above. I hope this clears
my take on the situation .

A couple of benchmark websites I tend to trust.

• Tom's Hardware: Hardware News, Tests and Reviews
• HiTech Legion

And obviously our own;

• PC Custom Builds and Overclocking - Windows 7 Help Forums
• Hardware & Devices - Windows 7 Help Forums