Hi Guys

Thanks for having a look at this.

I have my suspicions regarding this issue,but would prefer not to say what I think it might and see if someone else may have the same thought.

 

I currently have both p3dv2.2 and fsx running in dx 10 mode.

I recently upgraded my system to an i7 4930k oc to 4.5 an evga 780ti classified a rampage 4 extreme mobo and 16gb of ram.I also changed the casing and put a corsair 110 cpu water cooler in.about the only things that I left from my old setup was the psu (xpower pro gtx extreme 950w)and hard drive(Seagate)

Now the problem on the last few occasions I have run p3d I have the BSOD with various bug codes but not always the same one.

I am running the same scenario which is from EGLL to EGKK with orbx England orbx global and uk2000 egll and active sky next all are p3d compatable with there own setups. After about 10mins the system crashes to either a blue screen or just a complete lockup  requiring a restart.

I ran the exact same scenario in fsx  and all works fine no lockups or anything else.

I decided to try a few stress tests on my system to see if that might show a problem which it has but does not name it specifically but just says hardware failure.

I am stress testing using AIDA64 which was recommended by the people that supplied my parts as they also build systems and also OC my mobo/cpu combo.

Now this is where I found my issues.

If I stress test the cpu,fpu,cache and system memory  all tests fine

If I stress test the gpu on its own it tests fine also, but if I stress test all together then after between 8 and 15 minutes then it stops running and tells me I have a hardware failure, but I have no bsod or lockup

Now with this information can anyone suggest what it might be.

As I said I think I have an idea but hopefully someone will come up with the same thought. Don't want to go buying stuff that it might not be

 

Thanks

 

 

 

 

 

With any issue like this, the first step is to reset to UEFI defaults. In other words... no overclock.

Most likely is that your PC supplier hasn't provided you with a stable overclock.

Testing an overclock properly takes time. Often you find issues don't show up for weeks. Doubtful they would have tested as thoroughly as is recommended.

 

Most likely scenario is insufficient CPU voltage or RAM timings/voltage.

 

EDIT: I should add, if you decide to revert to optimised defaults, then make a note of the overclock settings, or save as a UEFI profile.

Yeah, probably an unstable OC. 

How did you go about OCing exactly? 

It was overclocked for him I believe Dazz.

It was overclocked for him I believe Dazz.

 

Oh, I assumed he built the rig himself because he mentioned he cannibalized the old PSU, but if that's the case maybe he should contact whoever overclocked it before resetting his BIOS in case it voids his warranty 

Hi Guys

Thanks

In regard to the OC the system is a bit of both.

I purchased a new mobo/cpu that was overclocked by the supplier also the mem

I also bought a new gpu and casing once it all arrived I assembled all the bits adding in my psu and hdd

I originally was suspecting that the psu might be the culprit but having done more testing I think the oc might not be as stable as advertised ( will be speaking to them tomorrow) I have a bit of headroom to increase the voltage as temps and core voltage are well under max but want to chk with them regarding warranty etc

The voltage is sitting at 1.35 and temps are 25-30c below tgmax

I believe I can take the core up to 1.4v but as its already stable for most day to day stuff I'm hoping moving it to 1.36 or 7 might be enough

Sounds like an unstable clock to me too. I've seen, on my machine, an AIDA or Prime95-test run for hours without a crash, only to go belly up when fsx.exe fired up the autogen..;) 

What are your temps? 

Are there any errors in:  "Event Viewer (Local)" expand "Applications and Services", then "Microsoft", then "Windows", then "Kernel-WHEA", and then finally "Errors" ? 

Is the memory overclocked?

In my case, I configured the EventLog to pop up an info-box every time the EventLogger logged a WHEA (Windows Hardware Error Architecture) -error related to CPU. More here: http://www.overclock.net/t/1317335/whea-error-alert-guide-or-how-i-got-out-of-wheaville

Prime or AIDA never triggered a WHEA-error while running tests, but FSX (or P3D) did - and in some cases still do. Seems like 100% on the CPU doing the same test in a loop is not the same as 100% doing autogen and what have you in FSX/P3D..;) I could never get my clock stable at more than 5.0 gHz (see my specs) with more than 2133 mhz (9-9-10 if I recall correctly) of memory clock (2600 ram) - the integrated memory controller just wouldn't take it. I traded the gHz for memory clock, as I couldn't prove that memory bandwith had a similar impact on performance as CPU-speed. Fair trade. 


Best of luck!

Regards from Norway, 
Kristoffer 

 

Good point on the memory OC.

 

Pete, post a couple of screenshots of CPU-Z with the Memory and SPD tabs or check yourself if it's running on spec

Hi guys

and many thanks

Kristoffer I just did a search for event viewer and from what I can see in the last 24 hours I have 25 errors in the last 24 hours !!!

however most of them seem to be either kernel event tracking or service control manager. unfortunately I know almost nothing with this side of computing so it could mean anything to me..

I also happen to have blue screen viewer  and with the blue screens that I have had the thing that has caused most of the faults seems to show ntoskrnl.exe as the cause but the bug chk codes are mostly all different. don't know if it means anything to you.

this is what that shows

 

062214-12573-01.dmp 22/06/2014 03:40:08 MEMORY_MANAGEMENT 0x0000001a 00000000`00005100 fffff6fc`40028958 00000000`00000001 80000003`b3494963 ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\062214-12573-01.dmp 12 15 7601 305,642 22/06/2014 03:41:13 
062114-20420-01.dmp 21/06/2014 03:09:47 KMODE_EXCEPTION_NOT_HANDLED 0x0000001e 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 tcpip.sys tcpip.sys+16d9a0     x64 ntoskrnl.exe+75b90    C:\Windows\Minidump\062114-20420-01.dmp 12 15 7601 305,514 21/06/2014 03:10:29 
062114-20248-01.dmp 21/06/2014 02:36:01  0x00000124 00000000`00000000 fffffa80`0ebef028 00000000`be200000 00000000`000b110a ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\062114-20248-01.dmp 12 15 7601 305,514 21/06/2014 02:38:24 
062014-20716-01.dmp 20/06/2014 22:27:42  0x00000124 00000000`00000000 fffffa80`0ddea028 00000000`be200000 00000000`000b110a ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\062014-20716-01.dmp 12 15 7601 305,514 20/06/2014 22:28:48 
062014-20436-01.dmp 20/06/2014 17:19:45  0x00000124 00000000`00000000 fffffa80`0ebd3028 00000000`be200000 00000000`000b110a ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\062014-20436-01.dmp 12 15 7601 305,514 20/06/2014 17:21:02 
062014-19765-01.dmp 20/06/2014 16:41:44 IRQL_NOT_LESS_OR_EQUAL 0x0000000a fffff080`037e2792 00000000`00000002 00000000`00000001 fffff800`05080eb0 ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\062014-19765-01.dmp 12 15 7601 305,514 20/06/2014 16:42:25 
061714-19515-01.dmp 17/06/2014 16:49:08 DRIVER_IRQL_NOT_LESS_OR_EQUAL 0x000000d1 00000000`00000002 00000000`00000002 00000000`00000008 00000000`00000002 ntoskrnl.exe ntoskrnl.exe+75bc0 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.1.7601.18409 (win7sp1_gdr.140303-2144) x64 ntoskrnl.exe+75bc0    C:\Windows\Minidump\061714-19515-01.dmp 12 15 7601 305,514 17/06/2014 16:49:48 
 

 

It doesn't copy and paste well unfortunately

As far as I'm aware the memory was not overclocked

Rgds

Pete

Just to add

I tried the kernel error through the WHEA and it shows no faults the fields are all blank

I'm wondering if this might be driver related??

Yeah, probably an unstable OC. 

How did you go about OCing exactly? 

 

 

As Dazz has stated my money is on OC - I had similar problems until I got a stable bios OC - event viewer is useless doesnet give you the real reason why it crashed just the result of the crash from what I have noticed - remove that OC and if that does not help you have a hardware issue - only way to figure that out is to remove that OC

 

4.6 OC is very high if your bios is not perfectly setup

Hi guys

Ok have been playing about a bit and discovered some interesting things

While I think you may be correct regarding the overclock ( I have to change my profile as its 4.5 not 4.6) as Ido occasional get an unstable msg running intel burn test

I don't get a msg while running Aida64 or prime 95.

Anyway having done a bit of reading I created a quick scenario with the default p3d2 and ASN start and the f22  with my antivirus on and off(Microsoft security essentials)

with AV ON and I take off rapidly climb to 30K+ roll over dive to ground with sharp turns and get to about 1000 ft climb again and repeat within 2-3 cycles of this my system crashes

But if I have AV OFF I can do this quite happily without any system crashes.I've been doing it for 15-20mins eventually overstressing the A/C but not my system

So while too early to tell I think my AV may be having a problem  with p3d

any thoughts?

Have you added p3d and all p3d related software in MSE excluded files and locations?

 

Should always be done.

Not yet

must do that.

 

I have spoken to the guy that oc my system.

We are going to have a test session tomorrow but a couple of thing he suggested might be down to temp related,even though all temps seem to be running within safe limits

But I tried running Aida 64 with stock settings and running all 5 stability tests and it runs fine for an hour then I stopped the test .so seems to be oc related but not sure yet if its down to voltage or temp.

 

Interestingly while I had the cpu at stock,I ran p3d and to be honest I noticed absolutely no difference in fps,admittedly not in a heavily built up area but still in the high 60+ fps range.

I would have thought that with the increased cycles ie 3.9 stock to my 4.5 oc I would have more fps

any idea why I'm not?

Thank to everyone who has helped so far with this as its been a steep learning curve

Not yet

must do that.

 

I have spoken to the guy that oc my system.

We are going to have a test session tomorrow but a couple of thing he suggested might be down to temp related,even though all temps seem to be running within safe limits

But I tried running Aida 64 with stock settings and running all 5 stability tests and it runs fine for an hour then I stopped the test .so seems to be oc related but not sure yet if its down to voltage or temp.

 

Interestingly while I had the cpu at stock,I ran p3d and to be honest I noticed absolutely no difference in fps,admittedly not in a heavily built up area but still in the high 60+ fps range.

I would have thought that with the increased cycles ie 3.9 stock to my 4.5 oc I would have more fps

any idea why I'm not?

Thank to everyone who has helped so far with this as its been a steep learning curve

 

 

Try 1.1 volts on cpu and a 4.2ghz OC - your high OC is not worth the aggravation 

The Ivy-E is strange CPU:s With your cooling i think you hit the roof, tested a 4960X in FSX had very hard to get it stable at 4.7 with custom VC.

It like cold its nothing to OC with Air or H110 you probaly end up in 4.2-4.4 range , the memcontroller is strong easy 2800mhz.

Not good performance in FSX poor IPC (single tread) at 4.7ghz it performed like a 4.2ghz 4770K.

That was the reason wy a stay with the Haswell.