Stability issues

krystan

krystan

krystan

Associate
Joined
4 Dec 2008
Posts
82
Location
leeds , west yorks
I recently completed my new computer build, I have built many other machines in the past and I got though this one pretty quick also.

I have the following spec

Asus Striker II Extreme
Intel Q6600 GO
2 x 2GB Crucial Ballistix (the ones with the led, sad i know)
Colled by asus silent knight II with mx-2 paste
I also have a soundblaster XFI fatal1ty titanium
And a BFG Tech 9800 GT OC 512 GDDR3 PCIe2
Samsung Spinpoint Sata II tb hd
1 x optiarc cd/dvd rw
All powered by an OCZ GameXStream Power Supply (700w version)
This all sits an an antec 900 II with all fans turned to maximum (side fan arriving tomorrow).

Everything looks cool (processor at 30 deg rising to no more than 38 under load) however i am suffering lockups At first I thought this was crysis so I have now run several benchmark tools (3dmark etc) and these also lock, and even world of warcraft seems to lock.

I am running nvidia 182.50 release WHQL drivers and have tried the new beta and old 7.x drivers and also get the same issue.

I have noticed that usually the locks will hit when my GPU hits about 57 degrees (it never hits 60) but that surely is in tolerance ?

I have made sure all cables are plugged in etc and am typing this on the pc I have built.

I am about to run memtest overnight to see if bad memory is causing this but seriously doubt it.

All i can think of is my PSU isn't powerful enough (doubtful) or the BFG card is faulty , or the nvidia drivers are trash.

I should also mention that all asus bios settings are set to auto and so I am not overclocking at all.

Any ideas ?

I'm going to include more info here about the processor just because I figure more info is good:

Processors Information
------------------------------------------------------------------------------------

Processor 1 (ID = 0)
Number of cores 4 (max 4)
Number of threads 4 (max 4)
Name Intel Core 2 Quad Q6600
Codename Kentsfield
Specification Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz
Package Socket 775 LGA (platform ID = 4h)
CPUID 6.F.B
Extended CPUID 6.F
Core Stepping G0
Technology 65 nm
Core Speed 2399.9 MHz (9.0 x 266.7 MHz)
Rated Bus speed 1066.6 MHz
Stock frequency 2400 MHz
Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, EM64T
L1 Data cache 4 x 32 KBytes, 8-way set associative, 64-byte line size
L1 Instruction cache 4 x 32 KBytes, 8-way set associative, 64-byte line size
L2 cache 2 x 4096 KBytes, 16-way set associative, 64-byte line size
FID/VID Control yes
FID range 6.0x - 9.0x
max VID 1.300 V
Features XD, VT
 
Last edited:
You may have a dodgey PSU, lock ups are often a symptom of this. Might be worth seeing if you can test it on another rig or test another PSU on this one somehow.

Memtest may reveal it's the memory, that's not impossible, though bluescreens tend to be more common with that (though hardly a rule!)

Go through the system bit by bit. My advice (failing the above) is always the same: remove all the non-essential components (i.e. all but 1 stick of RAM, then alternate) including any PCi cards, optical equipment etc. See if it passes DOS prime test (try prime test on ultimate boot cd) with everything unplugged, then work your way up by putting stuff back. Make sure you reseat RAM well.
 
I would have thought memory straight away if you haven't been using it for long.

Have you checked the memory is running at correct voltages and timings? Thats the first place I'd look.
 
The GPU should easily be able to run at those temps, not to mention the ambient in that case has to be low with all of those fans running at 100%.

When the PC locks, do you get any BSOD or error message at all? I assume if you did you would have listed it, but I have to ask :)

Also I would try running something like ATI tool or Furmark (both free and should heat up the card nice and good) to see what it maxes out at in those apps. You may find that the GPU has failed by either a lockup or BSOD in Furmark, or in ATI tool it will tell you if it's artifacting which is most commonly a failure in the video card. If that's the case I can take care of that for you.

It would also be helpful to see what your 12v rail is running at, that's a pretty decent PSU so I doubt that's the issue, but you never know.

Let us know when you get a chance.
 
as of this morning memtest revealed 0 errors after running several passes.

My PSU is a brand new game xtreme from ocz which was sent to me after another blew, I thought that a 700W gaming PSU should be enough to run this sort of system, especially as I am not using sli.
I also ran the furmark stability test for 45 mins and my gpu hit 67 degrees and kinda stayed there but no hang ups.

Driver issue ? Surelyif it was power it would have hung.

As for logs or BSODS of any description no I'm sorry the entire computer simply freezes up and I have to hit reset.

Also 12V rail running at 12.42v
 
Last edited:
There is still a chance of it being the PSU. The problem happens when gaming only? I don't think mainboard can be ruled out. Does anyone know if his mainboard has a temp sensor for the NB in case it's overheating?
 
The problem does run during gaming only seems to be absolutely fine doing anything else, I suppose the next thing to do is run quad versions of prime 95 and see if that hangs etc.

As we all know the only real way to test if this is the PSU is to spend about 100 quid on a new psu and hope.. which tbh is not my preferred option here, as thats an expensive way to test a psu. I don't really know anyone else running rigs like this so getting someone elses psu is not an option.

Unfortunately.

When I get home tonight I'm going to note memory timings etc, but I would have thought problems like that would have manifested during memtest overnight.

Anyone got any other ideas ? I'm open to anything right now. The NB temp is worth a look and i'll check that out (sensors present etc) and get back to you.

Thanks for your patience.. mine has been tested (bahh I wanted to play crysis :) )
 
Last edited:
ok i'm leaning towards ram right now. Not entirely sure what my timings should be etc so here is a little more info:

CPU Multiplier 9.0
LDT Frequency 5x
FSB(QDR) 1066
Mem (DDR) 1333.3 (is this right should it be 1066 to match the processor ?

timing: 6-6-6-20 1.8v (as written on the dimms)

I will try removing ram as StevenRaith has said, mainly because I just got a DRIVER_IRQ_NOT_LESS_OR_EQUAL BSOD when i ran the crucial scanner which says to me bad, memory or voltage wrong ?

Quick edit to add voltages as read out by my striker II extreme

CPU PLL 1.32
CPU VTT 1.5v
Memory 1.8v
NB Core 1.35v
SB Core 1.48v
DDR3 Termination 0.81v
3.3 rail: 3.31
5 rail: 4.94v
12v : 12.41v

--------------------------------------
This is the spec from crucial for my memory

4GB kit (2GBx2), Ballistix Tracer 240-pin DIMM (with LEDs), DDR3 PC3-10600 memory module
BL2KIT25664TA1336
Module Size: 4GB kit (2GBx2)
Package: Ballistix Tracer 240-pin DIMM (with LEDs)
Feature: DDR3 PC3-10600
Specs: DDR3 PC3-10600 • 6-6-6-20 • Unbuffered • NON-ECC • DDR3-1333 • 1.8V • 256Meg x 64 •
 
Last edited:
testing results based on advice given in this thread.

Test Results (Being updated as time goes by)

Memtest results

1) Now have well over 18 hours of continuous testing a the original timings (auto) with no errors at all

2) Testing at correct timing and voltage (as written on the sticker on the sticks): Passed no errors.

Video memory test(http://mikelab.kiev.ua/index_en.php?page=PROGRAMS/vmt_en)

Passed no lockups

Windows Based Video Testing

Tests done with no sound card and only 1 stick of ram. Ram was configured to the figures written on the stick with the correct voltage again as written on the stick.

Results (3d mark):

1) all combinations of ram (either stick) in all slots with video card in same slot all fail with a lockup with 3d mark, crysis and the video card stability test program they occur at random intervals, but it freezes then black screens on occasion meaning the reset button must be pressed. It is not heat related as its failed at 42, 57 and 60 degrees

2) All combinations of PCIE slots failed with same results as one

3) Went into bios and linked ram to fsb speed (made 0 difference still suffer lockups)


Results (Freestone stresstest)

1) Lockups and freezing here is a picture, this is similar to the corruption in games and in 3dmark vantage

Results (Furmark)

1) Never seems to lock up (this is open gl).



Dos based Prime Torture test

1) Full hardware Configuration: Passed after 9 hours no lockups.
2) Full hardware Configuration 2: Complete No Lockups
3) Minimial Hardware Config : TBA (10 hour test cycle).

Testing Of Video Memory using vmt (http://mikelab.kiev.ua/index_en.php?page=PROGRAMS/vmt_en)

Observations:

Only locks up when video card is under load in 3d applications
Hasn't locked up yet in an open gl application

I have had 1 BSOD from the crucial memory scanner but I understand this is a vista x64 issue with compatibility from that scanner and not my hardware (please correct me if I am wrong on this).


I am trying to approach this in a systematic way, thanks for your patience :)
 
Last edited:
OK.. I'm stumped and need help :mad:

To be honest guys I'm now at a loss as to what to conclude,

I can't see it being memory I've run a LOT of memory tests and all are clean.

My personal conclusion is its either the card or the psu (which is an ocz exteme gamer 700w psu which is SLI rated and I'm only running one card)... I suppose it could be that.

BFG have offered for me to RMA the card, I suppose i won't really know until I rma that and it either doesn't happen again, or i then need a new psu.

At that point I think we are looking at the board. CPU tests do not lock up and just continue and anything else that does not use direct x seems to be ok (will be running wow open gl and counterstrike later to doubly confirm that)/

It still happens with no sound card in so i've ruled out the creative stuff interfering.

I could really use a hand now I'm at the edge of my knowledge. :(
 
alexhull24 said:
Sounding a lot like the GFX to me. Is there anyone else who can test it for you in their system before you RMA?
Click to expand...

Not really unfortunately even though I work in a IT department no one wants to disturb their overclocked systems :)

There is one guy who is a good guy he has a corsair 500w psu tho not sure that'll support my card he is will tho to lend me a crappy old nvidia card of his sisters so we still have that option.

I was quite shocked BFG's first response was to rma :) did you see my picture of what was happening.

Thanks for the response I have never had such an issue with any other system i've built :(
 
TBH mate i'd try and get that card tested somehow. You could borrow an older card to try and see if it happens, would certainly rule out the card if it has the same fault, though you'd have to try something a little more powerful than budget if possible.

You could always try get hold of a card second hand like an old G80 nvidia GTS for about £34-£40 then sell it on when you're done with it. If it's not long I don't see why you'd lose any money, plus those cards eat plenty of power, so should test the system.
 
alexhull24 said:
TBH mate i'd try and get that card tested somehow. You could borrow an older card to try and see if it happens, would certainly rule out the card if it has the same fault, though you'd have to try something a little more powerful than budget if possible.

You could always try get hold of a card second hand like an old G80 nvidia GTS for about £34-£40 then sell it on when you're done with it. If it's not long I don't see why you'd lose any money, plus those cards eat plenty of power, so should test the system.
Click to expand...

Ok I've just bought a 9400GT to test with (1gb card).

Didn't cost much its not a premium card, but should do the job, i'll flog it on ebay later. I'll keep you guys informed
 
Continued information on this is that my 9800gt is not a standard 9800gt, it is from the OC range from BFG now I hadn't thought about this but is it possible because this card is overclocked albeit by the manufacturer that my PSU is not good enough to power this,

Precision reports the following:

Core Clock : 625
Shadow Clock: 1566
Memory Clock: 900
 
If your psu is working properly it's more than capable of running your enitre system with a clocked BFG 9800gt - tbh, it shouldn't even break into a major sweat.
 
Ta I'll wait for the test card then RMA this card to BFG found a large thread on NZONE which points to it perhaps being a card problem.
 
Quick update I have now got a 9400GT plugged into my mother board it ran all the way through 36mark vantage (albeit very very slowly) and no freezes.. same driver same mobo settings same slot.

This card draws its power from the motherboard and not from the pcie 6 pin so what I think I have shown here is its not a mobo issue in general.

BFG have given me an RMA number now and I'm going to have to use it *major disappointment having waited 2 years to upgrade my rig to something decent.

What do you guys reckon, still a possibility its the OCZ psu or is this now fairly conclusive it the 9800gt OC , I understand we'll never really know until such time i have a new BFG card here.

I don't suppose it could be the actual pcie 6 pin thats faulty.. I'm gonna run crysis (less prettily) to try and provoke a lock up but so far so good :)


[edit]

Crysis ran perfectly even when i overclocked the 9400GT (731 core, 1960 shaders and 515 memory) having to run it in medium (rather than the extremely sexy HIGH and VERY HIGH the 9800gt could run it at (then again whats the use if you only get 5 mins)

Wait for the rma i think hopefully i'll get sorted.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom