ORB Mother fight causes PC crash - Anistropic filtering the culprit?

[0_The_F00l]

Hello,

While fighting the ORB mother the PC crashes midfight (usually during the leg breaking phase) , PC completely shuts down ,

Turning off anisotropic filtering and limiting frames to 60 FPS allows me to finish the fight without crashes while still keeping the game running at mostly high settings.

This has been happening repeatedly.

Even running with medium to low with anisotropic filtering active causes crashes.

There are also a lot of performance issues i get - static frames , sudden drops in FPS , hiccups , mouse movement being jumpy and not smooth.

The crashes are happening when i am playing solo , running a Negating swarm Inaros (cause i dont want to bounce around like a beach ball for 80% of the fight) , Sobek , Pyrana , my Zaw and Imperator vandal.

MY PC Build:

AMD FX 6300

Radeon RX 480, Adrenaline 18.12.3 drivers.

8 GB RAM,

Windows 10

It is important to note that i can run a stable 120+ FPS on other missions (non open world) at high settings,

it is only during the Orb fight (and some times on the vallis when specifically hunting the bolarolas) that i have ever seen my PC crash so completely - no errors no BSOD , complete involuntary shutdown.

Assuming my Graphics processor is under tremendous load, but if it crashes even at low it needs to be looked into ,

i am assuming it is not simply a "you need a better card " problem.

If anyone has found any workaround other than turning Anisotropic filtering off , it would be greatly appreciated as well.

Edited January 6, 2019 by 0_The_F00l

[ArgusXen]

2 hours ago, 0_The_F00l said:

PC completely shuts down

...

static frames , sudden drops in FPS , hiccups , mouse movement being jumpy and not smooth.

You know, what you describe looks really similar to what happens when the computer is overheating... It's possible that, even if the rest of the game is running OK, this particular fight puts just enough strain on it to cause the overheat.

I don't know how good is the cooling on your motherboard, but I have the same CPU on a rather old motherboard, and when I just installed it, some elements of the motherboard tended to overheat, causing similar things to what you describe to happen in other games. Putting some coolers onto the motherboard's north bridge and VRMs eliminated this problem.

[0_The_F00l]

21 hours ago, ArgusXen said:

Putting some coolers onto the motherboard's north bridge and VRMs eliminated this problem.

I suspect you are right, it is rather old too. Might be time to bust out the old table fan. (I live in tropical area) 

[fjomand]

Hello fellow Tenno,

I'm having the same issue whenever fighting the profit taker orb. I've narrowed it down to the rockets impacting on my warframe or operator being the cause of my pc crash.

Now when the rockets are coming I run or look away and then I can do the fight without my pc crashing. 

I've tried medium/high settings. When I turn effects down to minimum (10) then it takes longer to happen. Also anisotropic filter off (or lower) causes it take longer before my pc complete crashes.

I've monitored my GPU during the fight and after it crashes by using logs. Nothing really out of the ordinary although I can see the rockets cause spikes in GPU usage and therefore 2 or 3°C temperature spikes. Also in total power drawn offcourse. Voltage remains constant through all.

I don't think it's a problem with too high temperature though. When I run stress tests on my gpu for an hour temperature goes almost 10°C higher then when I'm fighting the orb or normally playing warframe (on high settings).

Since I've been running this setup for more than 4 years I don't think it's my motherboard either. To make sure that doesn't overheat I've cleaned it with pressured air and everything is dust free.

I'm doing the fight with Chroma Prime - Tigris Prime - Tombfinger - Kripath

My setup is:

PSU: Corsair CX500 | Motherboard: Asus Z87-C | CPU: Intel Core i5-4570 | Memory: 16 GB DDR3 Corsair Vengeance | GFXcard: Asus Gefore 970 strix

I've turned down the clock on my gfxcard overall and set max power usage to 90% overall due to the known problem of the 970's sending a black screen out. That known problem hasn't been an issue for me since I've made that adjustment years ago.

I'm looking forward to other people their input or experiences if they have the same problem and if there's a solution for it.

 

[sam686]

1 hour ago, fjomand said:

When I turn effects down to minimum (10) then it takes longer to happen. Also anisotropic filter off (or lower) causes it take longer before my pc complete crashes.

Check more temperatures, like CPU temperatures and VRM (voltage regulator module) temperatures. (edit: I use Hwinfo64 to see more temperatures)

Try more stress testing programs like Prime95 and make sure it does not show any errors. Try some basic Furmark for graphics stress. Try both CPU and graphics stress testing at the same time to check if there is a problem with power supply unit not providing enough watts.

CPU-Z can show more information about the CPU. Is your CPU running at correct speed and not overclocked? Base clock, if changed from default, might cause stability problems with lots of stuff like PCI-E communication and DDR3 speed. These clock settings can be reset in Motherboard BIOS settings. To see motherboard settings, while turning on computer, press DEL or F2 a few times.

Edited January 31, 2019 by sam686

[fjomand]

Thank you for the input! I've installed Aida to monitor more of my temperatures and voltages. I'll keep you updated if anything special comes to light.

I'm not running anything overclocked. CPU and motherboard are completely stock.

Like I said only thing I did is underclock my GPU to avoid having the black screen problem that de Geforce 970's cards have. That was 4 years ago since I've done that and the same problem hasn't occured. Now it's complete system shutdown instead of it being black screen but computer and everything still running like I could still hear music playing in the background.

[fjomand]

Hi Tenno,

So I'm still not any wiser than before. I've been testing trying to log a crash using the HWInfo64 app but it doesn't want to crash anymore. Before I installed that app. I was running the profit taker orb with one of my friends on low settings (everything off or low or turned to minimum) and it crashed then as well. Wasn't logging offcourse silly me. After that I ran the mission again a couple of times solo on low settings. Coulnd't get it to crash. I got some reading though indicating that my GPU uses core 40% and 110 Watts on average during the 5 minute fight.

Why would it crash even when it's not on full load right? That's what bothers me most. It's just chilling doin't it's job at 40-50% capacity and sudenly it cut out.

So I got frustrated it didn't crash anymore (I know: lol whuut right? :D) So I did stress test on CPU for more than an hour. No crash a bit high temps (up to 83 °C) so I fine tuned my cpu fan a bit. 

Next I put a stress test on my GPU and my PC crashed after 58 mins. I go check the logs, nothing weird just everything on max power and stable voltages and temperatures (around 63 °C) VRM temp was at 90 °C though. I looked that up and found that that isn't anything out of the ordinary (It's allowed to go to max 110 °C). Anyway I ran the same GPU stress test again and it PC shut down after 10 minutes. Again nothing weird to find on the logs. That made me think it's my PSU that gives out even though I find that the air coming out of the back isn't really crazy hot. I don't have sensors on there so I don't really know whats happening in there.

Now I did the fight 5 times in a row on all high settings and nothing happend. GPU load on average just below 50% (during rockets 80%). Temperature never went above 53°C.

I have some voltage readings and when the rockets are landing +3.3V and Vcorerefin fluctuate 20 mV and 10 mV but I see those fluctuations happening frequently (not only during the rockets). So I think it's OK too.

Next thing I'm trying is swapping my gfxcard with a 970 (MSI instead of Asus) of my friend and run the mission to test that out. After that maybe try swapping PSU's.  If anything interesting happens I'll post it on here. 

[sam686]

8 hours ago, fjomand said:

So I did stress test on CPU for more than an hour. No crash ... Next I put a stress test on my GPU and my PC crashed after 58 mins. ... I ran the same GPU stress test again and it PC shut down after 10 minutes.

Crash, as in the computer just cut off power and windows log shows unexpected shut down errors? If yes, the problem is likely PSU or possibly graphics card. When things was working fine and now it randomly crash or power cut off, something may be dying, but which one?

Other possible cause of sudden power cut off in desktop computer is unstable electricity. If computer power cut off and lights flickers at the same time, poor electricity is likely the problem. Uninterruptible Power Supply will greatly reduce poor electricity problem. I use APC XS 1500 (BX1500G) in United States and have seen this click into battery mode for 1 second and lights flicker, this have helped my computer to stay on.

[fjomand]

On 2019-02-03 at 7:41 AM, sam686 said:

Crash, as in the computer just cut off power and windows log shows unexpected shut down errors? If yes, the problem is likely PSU or possibly graphics card. When things was working fine and now it randomly crash or power cut off, something may be dying, but which one?

It shuts down completely and reboots. Logs say critical error. Due to not responding or being stuck or because power cut out. I live in Belgium and our grid is supposed to be pretty solid. Like you said it's either my PSU or my GPU. I'm trying to get my friend over here with his computer to change GPU card and afterwards PSU to eliminate what it is exactly.

Crazy thing is today I was looking in Geforce Experience (crappy app I know) at recommended settings for warframe and I saw I had 2 of the settings in warframe lower than what it recommends. So I changed then to a higher level, ran the mission and checked logs. My GPU drew around 30 Watts more power than on the previous "default" high settings in warframe and got 4°C hotter. I didn't crash though.

This is turning in quite the adventure I know. Thanks for the advice!

[sam686]

1 hour ago, fjomand said:

It shuts down completely and reboots.

It looks more like a graphics card problem.

 

Your CPU (Intel-Core-i5-4570) have integrated graphics. With multi-monitors, use it by plugging one monitor into motherboard's video port, the other monitor into NVidia graphics. Might first require it to enable it on motherboard settings. when turning on computer, press Del or F2 multiple times to enter settings. Advanced menu, System Agent (SA) Configuration, Graphics Configuration, iGPU Multi-Monitor setting, Enable. Then save changes. Install intel graphics. Edit: go to windows monitor settings to make sure NVidia is primary.

Now, the reason for this is, if graphics card fails into black screen again, the other GPU will continue to send video output to see what is going on. Frozen / locked up? blue screen on death (BSOD)?

Edited February 4, 2019 by sam686