Some System errors show like this ACPI 2 one and cplspcon one, but I've seen these on other Windows builds at work also. Might reinstall Windows 10 one weekend to see if that helps though.
Appreciate the extra research done; still finding out extra things I didn't know about how Apex handles the 3000 series cards (or similar).
If you reduce the temp limit you also reduce the power draw of the card.
If a game crashes the whole system it is almost always a PSU ruining into the OPP and I saw 3090 crashing system with 1000w PSUs due to load peaks that are fare higher than the TDP.
That it does not happens with your 2080ti is one more hint that the problem is on the hardware side and not software related, there is almost nothing that could crash the entire system, without a blue screen, on the software side anyway.
So imho the PSU is still the most likely culprit here. (Or the power draw of the 3090, however you see it.) 😉
I would either cap the FPS in half or seriously under-clock the 3090 to see if the issue persist.
If not it is one more hint towards the PSU.
It is strange that the game freezes on an under-clocked 3090.
May I ask how much you under-clocked the card?
Otherwise, if it works for you it is fine. 👍
I just got this with completely different GPU, I have RX 6800 XT. Just got black screen and after that reboot.
I don't have any OC set up neither for CPU nor GPU. Just standard settings for both. I am using latest Adrenalin drivers.
The important thing to know is if your PC is shot down without a blue screen and then restarts, or if the system just hangs?
That is still not ideal.
Did you ever run stability tests on this system?
Or 3DMark, Time Spy, Firestrike?
I got this right, you system is just powering down?
Why do I 'feel like' this game is killing PC? We all have a PC to play this game but can't maximize the PC potential.
If we try to maximize the PC performance, the game is killing PC. Some might say "Why other games are ok but Apex is just killing it?"
I think the game just keeps getting better...like what happen to the new world killing 3090?
I'm sorry I'm not helping you guys. I don't have games to play as the game keeps disconnect me.
Darkness9t3 wrote:
Haven't tried 3DMark, Time Spy or Firestrike yet, should I try them all? Will start with 3DMark and let you know.
It can't hurt. 😉
Darkness9t3 wrote:
My next step though is to try an OS reinstall at some stage, and if it still happens I'll look into a more powerful PSU because the RTX 3090 does seem pretty power hungry.
Modern CPUs and GPUs have very effective protection mechanism against overheating, it is rare that a PC is just crashing because of overheating in our days.
As said, a power issue namely the PSU running in its overpower protection is fare more likely.
You do have the newest graphic driver installed?
It still could be something else, like the power supply of the Motherboard, so I would suggest to buy at a place that allows returns.
Your RM series is not a bad PSU at all, but for a 3090 I would recommand something in the HX/Platinum class.
So I don't have 3DMark Purchased but I just used the Demo Benchmark for Time Spy, here's the results. All ran smoothly though and no freezing / crashing issues.
Also Apex hasn't frozen again yet since using DDU after the latest Graphics driver, so it might just be a case of conflicting graphics drivers because I haven't usually used DDU between each update.
If it works, it works. 🙂
@Darkness93_yt Speccy is very useful and will allow you to monitor the temp in windows. Can put it into startup and into the bar to monitor constantly. Free download here.
https://www.ccleaner.com/speccy/download
On a practical level, I occasionally gently hoover the insides of the unplugged pc, the fans esp (put finger on fan to stop it spinning wildly) and vents, to stop dust building up, and make sure it's generally getting well ventilated.
@maximas1986 I don't think the latest crashes are to do with the @Darkness93_yt 's heat issues.
Yeah if you're talking about disconnects (there's server issues lately) or crashing to the menu / shutting down the whole PS5 that's a different story, I'm playing on PC so can't comment much on that.
@OldTreeCreeper Thanks but I'm familiar with Speccy, I use Geforce Experience's Advanced Performance Overlay instead, which monitors GPU temp as well as Fan Speed, Power, Voltage etc. It's what prompted me to play with the fan curve & temp limit initially as it seemed to crash on around 83 degrees at stock.
Just an update anyway, it happened again tonight (instant system power off) but I had only just used DDU this morning and installed the latest graphics driver to help prevent it. It might again be something to do with latest Nvidia drivers not quite being ready for the 3000 series, but as shaders are re-compiled and the game run a few more times that's probably what naturally makes it more stable.
Having said that, I never had these issues with my RTX 2080ti or my previous gaming rig (where the 2080ti normally lives). The game may have frozen sometimes, but not a whole system power down.
I'll see how it goes through the month but it seems to line up with a new Nvidia driver each time.
The weird thing is I can play other modern games on ultra too and they don't get the GPU running nearly as hot / crash / shutdown the system like Apex does, even if it is only a few times a month.
@Darkness93_yt it's a very odd issue. Youngtree has a dell that runs fine with a gtx1060, but we put extra ram in to be safe and run Windows from the c drive ssd, and everything else from other big drive or I installed a few smaller drives to spread the load, esp for big games programs.
Do you defrag with defraggler after an update? By any chance do you have the windows defragmenter set to automatic, possibly running in the background as you play?
Have you gone through startup to make sure everything you don't want on, is off?
Have you checked the heatsink is flush and seated properly?
Have you looked at your bios settings for temp?
Have you tried leaving the case open?
Have you tried to troubleshoot removing devices systematically in the rig, to see if one of them is drawing too much power?
My laptop runs at about 60 degrees.
We've run into an issue with the rendering from solid works being very slow, but I suspect it's a setting somewhere in the background causing it. But no temp increase.
Nice points about defragging, I don't and have never needed to worry about it, it's an NVME SSD though, I don't really want to reduce its lifetime.
I haven't had it crash again since the one time right after using DDU for the Graphics Driver, so I think it's what I was saying about Apex needing to catch up a bit after shaders are recompiled after each update.
As for systematically testing I've only swapped the GPU out, left the side panel off also but it's a combo of the GPU and PSU that has the game crash at a specific point.
Might not be much else I can do until I get a better PSU / GPU for the system, but for now I'm still able to get the most out of the rig I won and it does seem to be happening less often.
Just wanted to know if the devs tested with RTX 3090's themselves and saw similar issues, but when I tried a Support Chat a few months back I was just sent in circles on trying different troubleshooting steps that I already did.. they did mention the more users try with those cards the more data they'd be able to collect from the game, so it's kind of a wait-and-see situation too.
SSD do not fragment.
Defragmentation is a thing on mechanical hard drives only. In principle it means that data that needs to be read in order to load an application or just data you need loaded at the same time is scattered over the surface of a mechanical hard drive, forces the hard drive to move the read head to jump over to a new location frequently.
That costs time because the read head needs to be adjusted every time, and that slows the drive down.
SSD don't have mechanical components, so it does not matter if the data is stored at different physical locations, the access will always be at the same speed.
@holger1405 yes, don't defragment an ssd. Even with a ordinary HD it's not a good idea to do it too regularly. Mostly I would use it to defragment fragmented files only, maby after a big update, and a complete defragment maby every 2-3 months depending.
Have found the windows defragmenter set to auto, on a few laptops with ssd. Think it's a default setting. Default for ssd should be off!
Yeah that's what I thought, haven't done any manual defragging for SSDs.
For "Optimising" (which I think means defragging) from the Windows settings though they're on defaults, should I just turn it off for the C drive and see if it helps?
Was also thinking to try not changing the temp limit on Afterburner launch if it's more to do with a driver cleanup after an update.
Darkness9t3 wrote:
Yeah that's what I thought, haven't done any manual defragging for SSDs.
For "Optimising" (which I think means defragging) from the Windows settings though they're on defaults, should I just turn it off for the C drive and see if it helps?
Was also thinking to try not changing the temp limit on Afterburner launch if it's more to do with a driver cleanup after an update.
No, Windows is detecting a SSD and will handle it accordantly, meaning it will do what is right for an SSD.
I said in my last post the SSD don't fragment, but that is not entirely true, even on a SSD it can happen that there are so many fragments on the drive that the Metadata, the data that tells the OS where a specific file is, can hit the limit and that would lead to I/O errors from the SSD.
So let the Optimising on.
Thanks for the info, will leave all unchanged then 🙂
So back to the original issue; I did some testing over this week and didn't change the temp limit at all.
Game still ran fine, which leads me to suspect it's something about the Driver Updates, and when the shaders are rebuilt for the GPU after one, sometimes it will just fail on a few games soon after but then stabalise itself.
There's currently another pending version (493.13, I'm on the previous one; 472.12, not sure why there's such a huge jump in version numbers though) but I'm skeptical to install it because I know Apex will crash out again a few times.
Any tricks with the shader cache I could try maybe? It's going to be something about the shader cache or the driver update, or maybe I should always stay one version behind (manually download it instead of downloading the latest with Experience)?
The big jump in version number is probably due to NVIDIA ending support for Kepler GPUs.
For the first time the new driver let you set the shader cache size manually.
But I really doubt that a shader cache problem could lead to a complete shut down of the system.
@holger1405 @Darkness93_yt I don't understand how not updating drivers is a good idea, as newer drivers usually synergise with the games better and it's always recommended to have everything up to date due to stability?
Yeah normally I'm all for keeping it on the latest driver, but I just have a feeling it'll make Apex crash again; it's always been in line with a GPU update it seems.
@holger1405 I doubt it also but it's a really weird issue, just my bad luck 😛 I'll do the update then I guess and report back after some more gameplays.
Question is; should I use my normal approach of DDU in Safe Mode or just go ahead with an Express or Clean Install? Usually I've used DDU, because when I go Express it's failed anyway. The Clean Install I've done every now and then but it still fails, so I'm not sure which is the most fail-safe approach.
Community Manager
