Graphic Cards Random Crash on a 7850 GPU (Asus 7850 2gb DC2 V2) [Urgent]

Techie#1

Techie#1

Skilled
Built a New AMD based system,
all components NEW,..except the Chasis/cabinet,periperials(keyboard/mice) & the PSU(Corsair VX 550w PSU) taken from from my older 5 year+ old rig

Now the problem:
# My Brand new purchased Asus 7850 2gb GPU card( HD7850-DC2-2GD5-V2 - Graphics Cards - ASUS ), on my new machine (though it works OK with the standard windows 7Ultimate 64 bit driver-it however is very slow , any work lags much & there is No Windows AERO transparency Effect too) but once i install the specific driver for the asus 7850 gpu & reboot...my machine randomly crashes & again reboots itself ( NO BSOD,No hanging or freezing of desktop). Even on windows login, the machine Crashes & repeats the reboot! My whole machine thereby becomes very unstable until i login through Windows7 safe mode & delete the drivers! After that, windows doesNOT crash anymore & become stable again but at the cost of BIG performance Hit.The Only thing I know that,..this is just NOT right & NOT done...I didNOT pay so much for the 7850 to receive this pathetic experience!

## This things happen even i install the drivers from the Asus GPU Driver CD that came along with the Asus 7850 , or install the latest Asus specific drivers for the 7850 from the Asus website or even if i install the latest 13.4 drivers for the AMD ATI 7800 series from AMD website.The system also crashes even when i just install "only" the latest driver for the 7850 gpu without installing CCC....there is NO difference in the post driver install scenario..system crashes begins to happen again!

### the system crashes occurs even on windows8 pro exactly similar to how it happens in win7 (did clean install of windows7 twice for resolving the case,but it doesnot help)

#### the system crashes with either of my two ddr3 ram sticks(Corsair Vengeance 8gb -4x2gb(CMZ4GX3M1A1600C9)

##### the crashes occurs even when the 7850 gpu is installed in the second PCI-E 16X slot !

##### the system crash continues even after updating my motherboard(ASUS M5A97 EVO R2.0)'s BIOS to latest one(from revision 1503 to rev 1708)

Further updates:
*Checked the card at local digicomp/asus service centre ,however they found the card to be good without crash or any fail..they played games,etc for 1-2 hours & no problems were found & so they didn't do rma & returned my gpu card!

** ran furmark on the 7850 & it ran good(the only time when the system didnot crash soon & could complete the test)

** i swapped the 7850 in my new system with my older XFX6800gs gfx card & then my new system runs perfectly stable & good with the XFX 6800GS without any sort of crash.Played some older games at highest ultra setting for 3-4 hours & everything ran great without any hiccups anywhere on my new system with the old gpu!

*** This is my first experience with ATI Radeon & had come from Nvidia platform where i never received such an ordeal!!! Now wth is this happening??? I am completely frustrated with the experience!
Is it because of unstable driver issue for the AMD/ATI 7850GPU or is there some hardware level problem with my New mobo or CPU or incompatible RAM (Corsair Vengeance 8gb -4x2gb(CMZ4GX3M1A1600C9) Ram or simply my old trusty Corsair VX550w PSU cannot drive the new GPU(7850) anymore?

**** In the Windows 7 "Event Viewer", I am finding this critical Event41, Kernel power errors, which directly corresponds to the random Window crashes which I am experiencing !
Look at the snap of Event viewer:
9vkprm.jpg


I really want to single out the culprit causing such harassment & resolving the crisis of installing the correct drivers of the 7850 & making it fully operational ASAP !

My new system configuration:
1)AMD FX8350
2)ASUS M5A97 EVO R2.0
3)ASUS 7850 2GB DC2 (HD7850-DC2-2GD5-V2)
4)CORSAIR VENGEANCE 1600FSB RAM ( 4X2=8GB)
* PSU : Corsair VX550W(2009 bought,.. from my old/current pc)
** HDD is a WD15EARS Green HDD brought in July 2010 (never used in any download rig...hardly used)!​
If anybody has any correct knowledge & similar experience like this,please do Help !​
 
Just one question : Did digicomp guys tested the gpu in front of you ? [ asking this because some one I know had to buy a new pair of ram unnecessarily for the clumsiness of the SC people ].

ANd another person I know had faced a lot of trouble for a faulty PSU which seemed like a memory issue and I myself faced lots of issue with a faulty cpu [ some games would run without any issue and some other's don't which looked like a gpu issue for couple of months ] - so make sure they are all ok.

For a quick try : There's a bios update [ Version 1110 ] available for your gfx card : try it .. if possible perform it on a friend's pc or from DOS mode.
 
Yes, Digicomp ppls ran game in front of me !

The only weak links in this case as I can find are:
1) the ram Corsair Vengeance 8gb -4x2gb(CMZ4GX3M1A1600C9) which is NOT listed in the QVL of the Asus Me5A97 Evo R2.0 motherboard. // yes I made an unfortunate mistake of ignoring mobos memory qvl.however I also honestly don't know if that matters & creates the system crashes// * Ran memtest86+ on both the ram modules for about 3 hours & no errors were found.

2)the PSU Corsair VX550 is 5-6 years old & had been running in a 24x7 download rig constantly since then. & so the aging capacitors of the PSU "MIGHT" not be continuously supplying clean power to the new system hence it might crash randomly.

However that being said, do notice even with these weak links in place, the system runs perfectly fine with a old Gpu like the 6800GS, but not with the new 7850 gpu.Now if the system runs perfectly stable in the above said configuration(procy, mobo & ram) but without the 7850 gpu, is it not logical to deduce that there is NO problem with the above components(procy, mobo & ram) except the GPU or the PSU or even the memory (outside chances) which may play the spoilsport!

My simple mind can think out about these only, though I'm utterly frustrated at playing detective with my new system anymore!
 
Crazy_Eddy said:
Your 6800GS has half the power draw of your 7850. What PSU did the digicomp people use?
Click to expand...
do you mean that at idle or at load?
though i don't know precisely what's the exact powerdraw of a 6800gs & that of the 7850 ,( i however used various online power calculator before purchase of the core components to find if my psu vx550 can supply the powerdraw of the system & i found it could very well do so i.e if my vx550 is still performing as good as in day1) but when the crashes occurs the exact feeling/thinking which comes instinctively into my mind is as if my 7850 is NOT getting the required power it needs to perform normally,even at idle i.e just at being in desktop.I think that when the 7850 is recognized by default in windows7 as just a standard vga adapter , its NOT opening its full potential (working much below its memory clock & Engine clock as i found in GPU-Z & AIDA-64) & hence is NOT drawing its full power requirement from the psu & so is behaving well within normalcy & hence no crash !...but once i install proper specific driver for the 7850 which loads on startup the 7850 is fully recognized by the system,operates at its full potential & wants to get its higher required power from the psu which it cant & hence crashes the whole system.The Asus 7850 GPU mentions it boldly on the GPU box that it requires a minimum of 550w of a PSU! This is one of the very simplified ideas which i can put up to reason with the random system crashes.
//This theory is however challenged by the fact that the 7850 gpu can run in its full potential & stable by a mere paltry mercury 450w psu at the digicomp service centre,though the other components there were less demanding //

Other idea is my ram Corsair Vengeance 8gb -2x4gb(CMZ4GX3M1A1600C9) is simply NOT compatible (as its NOT in my mobo's memory qvl) with my system & hence it crashes // the contradiction is my system does run perfectly good & stable with all the new purchased components (processor,motherboard & ram) with the old nvidia 6800gs gpu & hence the memory compatibility may NOT be any part at all in the system crashes //

topgear said:
mercury 450w.
Click to expand...
precisely +1, (costing rs 500) !
 
Power draw at load(why would I mention idle power?!) for the 6800GS is supposed to be about 70W. The 7850 is 130W. Surprising that its running off a Mercury 450W. I hope you're using the PCI-e power connector on your VX550 and not using any molex to PCI-e adapter.

Techie#1 said:
Other idea is my ram Corsair Vengeance 8gb -2x4gb(CMZ4GX3M1A1600C9) is simply NOT compatible (as its NOT in my mobo's memory qvl) with my system & hence it crashes // the contradiction is my system does run perfectly good & stable with all the new purchased components (processor,motherboard & ram) with the old nvidia 6800gs gpu & hence the memory compatibility may NOT be any part at all in the system crashes //
Click to expand...
The QVL is just for reference, people have been running local brands since the pentium I days. But instead of theorising so much, just run a couple of passes of Memtest86 and report back if it completes?
 
Crazy_Eddy said:
Power draw at load(why would I mention idle power?!) for the 6800GS is supposed to be about 70W. The 7850 is 130W. Surprising that its running off a Mercury 450W. I hope you're using the PCI-e power connector on your VX550 and not using any molex to PCI-e adapter.
Click to expand...
Bro..Honestly don't take me so stupid as NOT to connect the 6 pin PCI-E power connector ((the corsair vx550w comes with one of those straight out of the box & I'm definitely using that one only)) & report my issue as a problem to esteemed forum members here. If i don't connect the 6-pin pcie power connector to the GPU, & power on the system,the gpu would glow a Red led & fail from booting.

Anyway,here are the photos of the test rig & components the local digicomp guys were using to test my gpu:
1) My Asus 7850 2gb running properly/stable at the local Digicomp service center ( on their test rig,NOT Mine):

2) Details of the Asus 7850 2gb as seen at the S.C

3)The S.C machine on which my 7850 was running

4) The RAM on which the S.C test rig was running

5)The Mercury 450w PSU on which the S.C test rig was running to test my Asus 7850gpu !

* click on thumbnails for bigger picture
* * my mistake/carelessness that I didn't take snaps of game play on my 7850 gpu at the S.C.
***All total,I was about 20-30 minutes at the s.c before the machine on which my 7850 was running in front of me & games being played,but within that time i atleast didnot find any crash happening & the s.c guy operating the machine said he was up running the p.c & plying games for the last 2-3 hours & he didnot find any such trouble with his machine or running the 7850 gpu as i reported to them & i had no reason not to believe him considering that the machine seemed very stable in comparison to situation on my system where crashes seem to happen every now & then!

The QVL is just for reference, people have been running local brands since the pentium I days. But instead of theorising so much, just run a couple of passes of Memtest86 and report back if it completes?
Click to expand...
I even know that QVL is just for reference, but when you are clueless about troubleshooting the problem you tend to look everywhere & if you particularly look above at post no #3, ,i already said " Ran memtest86+ on both the ram modules for about 3 hours & no errors were found" & I'm NOT just theorising wildly...
Find here :


All the problem that shows up are a contradiction in itself...so i'm not being able to get any clue of whats really causing the problem! All this is just cracking up my brain & hence asking the experts help here!
 
ok, but cant you guess anything from the event viewer as posted above(in post#1)...there are some critical " kernel-power" errors logged in which is exactly when the crashes happens!...can they be power/psu related?
 
Techie#1 said:
there are some critical " kernel-power" errors logged in which is exactly when the crashes happens!...can they be power/psu related?
Click to expand...

http://support.microsoft.com/kb/2028504

When a computer that is running Windows starts, a check is performed to determine whether the computer was shut down cleanly. If the computer was not shut down cleanly, a Kernel Power Event 41 message is generated.

An event 41 is used to report that something unexpected happened that prevented Windows from shutting down correctly. There may be insufficient information to explicitly define what happened.

To determine what may have happened and to identify a potential resolution, it is important to know what the computer was doing at the time just before the event occurred.
Click to expand...
The kernel event 41 error is generated *after* the system has crashed and windows reboots. Hence the request to check what the BSOD message says *during* the crash. It could very well be a driver clashing issue as mentioned above.
 
I've faced such Kernel Power issue with faulty cpu .. like I said before half of the games would run flawlessly and others don't cpu was undervolted a little from 1.39v stock to 1.375v ] though fixed it lter by applying stock cpu voltage later but then again the faulty cpu just took 3 mobos with it .. so if possible check the mobo+gpu combo with another PSU and cpu .. I think Op is nearly at the brink to solve the issue so don't just give up now.
 
Two points contradicting CPU fault theory:
1) if cpu is faulty & generating the crashes,...then wouldn't it be logical to infer that similar crashes would been replicated when i put on the same machine(including the same CPU) but with a different gpu/vga card i.e the 6800gs but then i find NO system crash happen at all with the swapped gpu! I checked the cpu ,no bent pins,,nothing out of normality!

2) also I don't know if this is good enough check done by the motherboard itself at the booting stage when my motherboard Asus M5A97 Evo R2.0 checks for component errors & failures(including CPU,RAM,VGA Card & HDD),& gives a LED status for each of them...look below(as taken from the Asus M5A97 EVO R2.0 manual) :
onbord_led_check_m5a97evor2_0.png

* in my case ,None of the LED stays lit up atleast & the booting proceeds on
** i however don't how much precisely they can detect cpu failures

In anyway case if there is indeed any cpu failure which is NOT apparent, then how do you(reasons) /on what ground do you claim replacement?

*** 1 more thing to add ,which maybe quite imp for troubleshooting..is that...during the boot phase when the screen is momentarily blank/dark, i momentarily(maybe split of a second) see a green coloured mesh of horizontal & vertical lines as in a mesh ..dunno if it indicates anything !
 
^^ I'm only suspecting the cpu but it could be anything like psu or mobo though unlikely I know someone whi got 2 doa boards but no.3 worked ok for him

about cpu rma : you need to send a mail to amd requesting rma.. once you get the number you need to take a print out of the approval e-mail and deposit it the local service center.
 
topgear said:
^^ I'm only suspecting the cpu but it could be anything like psu or mobo though unlikely I know someone whi got 2 doa boards but no.3 worked ok for him

about cpu rma : you need to send a mail to amd requesting rma.. once you get the number you need to take a print out of the approval e-mail and deposit it the local service center.
Click to expand...

but don't the normal/stable running of the system with the 6800 gs provide any indication that the rest of the components are ok, or is that the cpu/mobo/memory faults will occur only when i install the 7850gpu?

& what about the green mesh that i momentarily get while booting...does that or not indicate anything about the o.s/system/gpu or anything out of place?

***Also very important, even if i may login to windows & desktop & be temporarily stable for sometime, but running system rating (WEI) definitely/sure shot crashes the system at any point of running the WEI !
 
can you post the crash dump here? check for *.dmp files in the system. also enable the dumps from startup and recovery settings.
 
Minidump:
.dmp file here > http://www2.zippyshare.com/v/37506660/file.html
here's a screenshot from "whocrashed"
20130729_153508.png

currently on win 8 pro !

but the strange part is the time stamp doesnot match the time of crash ,dont know if "whocrashed" is getting it correct. Only Windows7 System Event Viewer is getting the time correct when the crashes are exactly happening & generating Critical Error Event 41 code as seen in post#1!

Also strange is I dont get get any BSOD even after disabling "Automatic restart" at Startup & Recovery...while system crashes(sudden BLACK screen),it just reboots itself without any BSOD whatsover (even after disabling auto restart)!

just going crazy on the symptoms & eternal failures ! :mad::sour:

 
Back
Top