• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

PC shuts off when using graphics intensive apps

A

Adidas4275

Supreme [H]ardness
Joined
Mar 11, 2006
Messages
5,986
Specs:
3930k
Asus P9X79 Pro LGA 2011
Corsair 1333 32gb 4x8gb RAM 1.5v
R9 290x or R9 290
Corsair HX850 PSU
Corsair H100i 4x120 push pull
Fractal Designs Midi R2
Windows 8.1
intel 520 240gb SSD
Blu-Ray Drives


Okay so I am running everything on stock settings and it is stable, I can run Prime95 for 24hrs, or intel burn in or OCCT without issue, CPU 100% the whole time, temps are around 60-65*F depending on ambient (it is over 105*F here in the summer and the house varies depending on if my wife is home or not :)


The strange this is when i play Bf4 or run futuremark the computer shuts off, suddenly to black screen and will not turn off unless I power cycle the PSU.

I have a R9 290 with Gelid cooler and the R9 290x tri-x and temps are below 65 on those the whole time...

The PSU is new, the CPU,GPUs and Motherboard are used. The ram is 2 years old but was purchased new.


my first instinct is the PSU is bad (even though it is high quality and new)

so i ordered a PSU tester and it will be here in a couple days, i dont have access to another PSU that I can run the R9 290 off of.


I dont think it is the GPU since i tried 2 of them.

I RMAed my RAM because the windows version of MemTest threw out errors...

I pulled all but 1 stick today and was gaming on it and it still shut off.... then i ran the dos version of memtest without errors... so I am unclear on the ram.... new ram will be here Tuesday so I will test it.


Anyone have any other ideas? I have loaded default CMOS as well as removed the battery and let it sit unplugged for an hour or so to clear it...


Specifically is there anything i can look for with monitoring software to determine if this is a PSU issue? Thanks!
 
Last edited:
Quality brand PSUs are not totally infallible. What you have described does sound like a PSU issue to me - the 290/290X does draw a lot more power than the CPU at full load, which could explain why only graphical loads kill the PC.
 
thanks, i just ran OCCT on PSU mode and it shut off again with the 290x installed.

With an older passive 5550 card it runs for 30+ mins without issue.

Looking at my power draw off my UPS or the kill-a-watt I am around 500w so the 850w PSU should be enough...

Neweggs PSU calc says I need a 1200w PSU but the extreme power supply Calc shows 650w

I think the 850 should be enough.....

I filled out the RMA from corsair and will get it later in the week.

thanks
 
OP,

While I reading your issue I felt like I was having a flash back to an issue with my sig rig..Went through the same kinda crap, of course everyone screamed it was my PSU from the heavens despite the PSU having powered a platinium O/C'ing 7950 (1.325Ghz core/1750Mhz Mem) with the same parts aside from my 290 and an enlarged loop (added a second rad)...

I could Prime/IBT/Furmark for 24 hours, but if I tried to run the two together, BAM, instant hard power off...I tried the OCCT "PowerSupply" Tester and that app is junk..I ended up pulling my PSU and putting it on a load tester and a scope and everything was fine, yet if I trusted that "PSU Tester" I would have wasted time RMA'ing my wonderful X850 back to SS for nothing..

The issue I found, was my first PCI-E X16 slot..I moved the 290 to the second slot (which was not a small feat considering it is a custom loop) and everything was butter..I was about 10 hours from RMA'ing my 290 (tested it on a Sunday night, already had my RMA open with Sapphire) and would have ended up giving up another well above average O/C'er (1.275Ghz core/1375Memory)..

I spent 10 mintues on Google and ran into several posts on various forums where others had the same issues with the Z77 MPower and the first PCI-E slot being bad..Funny I never found them in the 5 or 6 hours of troubleshooting tips before:rolleyes::rolleyes:..
 
At stock, you shouldn't even be close to pushing that PSU, and bad power generally causes instability, not complete shutdowns. I'm more inclined to think your problem lays with either your motherboard, GPUs, CPU, or RAM. In my experience, the most likely culprit for hardware failure has always been the motherboard, followed by the RAM, then the GPUs.
 
Thanks guys for the help.... staying up late reinstalling W7 on my HTPC so my wife doesnt have to deal with no TV tomorrow....

So, I did Corsairs RMA for the RMA and PSU since the RAM gave MemTest errors in windows and they are 2 years old.

The PSU will be a fairly simple swap since it is mostly modular...

If I swap those parts and I am still having shutdowns I will go to the motherboard next.

Bought the CPU and Motherboard off the BST forum here a week or so ago so we will see how an RMA goes with a 2nd hand motherboard (Asus)

I hope for simplicity that replacing the RMA and PSU fix the issue.

I dont think it is the GPU since i have used 3 of them with the same results

It is completely possible that the seller never new the motherboard was faulty (if it turns out to be) since you would have to push the board hard with a highend GPU and it is possible he didnt have one.

My backup card is in the rig now and it works perfect even with running Prime95....

Again thanks for the suggestions. I dont mind trouble shooting and the $$ saved buying used gear is made up with time :)
 
The major motherboard manufacturers go by date of manufacture, not by receipt. So that should be relatively painless.
 
Those 290's will shut everything down if they are overheating. I just learned this today while attempting to reapply thermal paste to a 290... :eek:
 
Thanks. I'll run my HSF at 100% and see if that changes anything. I have a tri-x with a huge HSF with 3 fans and temps are not over 55C

But I'll try it for sure


I moved PCIe slots with no improvement

I swapped PCIe power cords to the alternate ones... No improvement

I disabled turbo on the CPU which decreased power draw and hear... No improvement


I can run Prime95 all day with out a shut down, but any GPU stressing program and it shuts down

Used 3 GPUs with the same issue

So I am thinking PSU or Motherboard
 
I hope. That is an easier fix and why I posted in this forum.

Thanks for the reply.

I'll definitely post here when I get it sorted out
 
okay, so early on in testing but i swapped it out for a 2008 built Antec Earthwatt 650w PSU with 3 12v rails capable of 25A each. 1 PCIe 8 pin on 12v1 and then a 2 4 pin molex to PCIe 6 Pin on 12v2

so fair it runs OCCT fine, the Corsair hx850w did not, and now i am going to game for a little in BF4 on it.

funny that this 640w PSU can handle this....
 
Then the HX850 most likely had a defective component. Probably something on the 12v rail failed, causing the voltage to drop as current was increased. Once it dropped past a certain point, protections kicked in and shut it down. You were probably in the unlucky 0.1% that gets defective high quality power supplies.
 
yeah seems like it. it was one of the only items i bought new and not used :)

that is okay, my old antec is getting the job done right now. I just put in about an hour of gaming BF4 eyefinity and it was great...

i did turn off turbo on the CPU and it runs a lot slower but cooler and doesnt make any difference in BF4.... so that was interesting to find out.

also i have 1 stick of ram (1x8gb) in to test it and gaming it doesnt make any difference either over a quad channel kit..

So really GPU limited... interesting

thanks guys for helping me through this. Hopefully ill get a shipment from corsair late next week with a new/working PSU
 
BF4 and most other FPS games are primarily GPU bound. It is only in large 64 player maps does BF4 become CPU bound.
 
yeah, i really only play 32-64 MP servers... but I am still seeing 40-60% CPU usage... the 3930k is a fast CPU, i think I saw that there is no difference between LGA2011 CPUs and the Devils Canyon CPUs interms of gaming today


Anyways, it is looking good and I do have a PSU tester coming tomorrow, so i will test it and make sure that is the point of failure.
 
update, Corsair says the HX850 is backordered and it will be 2-3 weeks before they can replace it.

hmm...

does anyone know if the AX series or another series uses the same modular cables as the HX?

Thinking I could buy a fully modular one and keep 99% of my wiring in the same place, only needing to remove the PC 8 pin, 24 Pin and the PCIe cables... leaving the SATA and 4 pin molex....
 
Almost assuredly not. They are probably not made by the same OEM even. You would have to ask someone at Corsair possibly.
 
A PSU tester will not tell you anything if your problems are under load. You will need the $3k+ equipment [H] and JonnyGuru have.

No, HX cables will not work with AX cables. To throw even more of a wrinkle into that, AX i cables will not work with AX PSUs, and vice versa. And AX1200 cables will not work with AXx50 power supplies, and vice versa.
 
OP
I've got the HX850 myself and its been bullet proof for me but alas sometimes things do break down.
I did plug your components into PSU calculator and you're well past 700. If you aren't merely replacing the HX850 but actually getting a new PSU I'd frankly look for something over 1K watt just so you have some margin
 
GonzoP said:
OP
I've got the HX850 myself and its been bullet proof for me but alas sometimes things do break down.
I did plug your components into PSU calculator and you're well past 700. If you aren't merely replacing the HX850 but actually getting a new PSU I'd frankly look for something over 1K watt just so you have some margin
Click to expand...

PSU calculators are known for overestimating needs.
 
Bluesun311 said:
Almost assuredly not. They are probably not made by the same OEM even. You would have to ask someone at Corsair possibly.
Click to expand...

okay, thanks... that is a bummer

Tsumi said:
A PSU tester will not tell you anything if your problems are under load. You will need the $3k+ equipment [H] and JonnyGuru have.

No, HX cables will not work with AX cables. To throw even more of a wrinkle into that, AX i cables will not work with AX PSUs, and vice versa. And AX1200 cables will not work with AXx50 power supplies, and vice versa.
Click to expand...

man that is really lame about corsairs modular cables... and that makes sense about the PSU tester

GonzoP said:
OP
I've got the HX850 myself and its been bullet proof for me but alas sometimes things do break down.
I did plug your components into PSU calculator and you're well past 700. If you aren't merely replacing the HX850 but actually getting a new PSU I'd frankly look for something over 1K watt just so you have some margin
Click to expand...

Yeah Corsair is well respected and I am sure I just got a bad one.

I have a a kill-a-watt as well as a UPS and both read under 650w OCed and fulled loaded running PCmark and Futuremark...

But I did just get a AX1200 off the BST to use since Corsair cannot replace the HX850 for 2-3 weeks + shipping time. :(

So perhaps that was a good thing, gives me overhead to add another 290x someday

Thanks guys, my nice wiring job..... for nothing :mad::rolleyes:
 
You must log in or register to reply here.
Back
Top