Goodbye ED... or is it? CPU activity and thermal shutdown issues

In b4 "get a new computer" which absolutely IS the right answer. I can't afford a new one though [sadz].

My computer's pretty old. FX-8350 [stock cooler], 16GB RAM, RX 570. Runs Ubuntu. I've been playing ED for a few months, 200 hours or so. Not many. I noticed relatively early on that I was getting thermal shutdowns occasionally, esp around starports. I put this down to the CPU overheating because of the stock cooler rather than the GPU or the 550W PSU. I also think it's the CPU because the PC shutdown the other day just when processing the Vulkan shaders on Steam before launching the game!

I have put all graphics setting to low, to no avail.

After a few ED updates I've noticed the shutdowns happening much more often. I literally cannot play the game now for more than 15 minutes without the PC shutting down. This makes it unplayable on my system as far as I am concerned.

Since I can't get a new system, what does anything think the LEAST CPU-intensive parts of gameplay are? Being in and around starports is a no-no. But I've noticed that in far-out systems everything is fine.

Is there anything I should do or anywhere I can go where I can hang out for a longer amount of time without the thing shutting down?
 
Interesting. So I had real CPU problems too (i5 4690k), especially after one of the more recent updates. What happened in my case was that the CPU usage for all 4 cores was flat lining at 100% with Odyssey running leaving nothing for other background processes that I needed like the joystick curves and head tracking software. What I had to do was specify CPU affinity (using a utility called Process Lasso), telling Odyssey to use CPU's 0-2 and my other bits of software to use CPU 3. Anyway, that kinda worked (barely) but eventually I was fortunate enough to be able to get myself a new PC.

Not sure any of this helps you much but you could at least try setting CPU affinity to tell Odyssey to stay off some your cores (which could at least theoretically help)?

The other reason I chipped in here was to say that while I was going through all this I learned that turning your graphics settings down can actually add to the load on the CPU (because the GPU is no longer the bottleneck) so there is a chance that turning graphics settings up might actually make your CPU less likely to overheat.
 
Oh, and to answer your original question, one answer would be to switch back to playing legacy 3.8 which is a lot less CPU/GPU intensive than live Odyssey 4.0. I don't blame you at all tho' if that's not an acceptable answer.
 
Interesting. So I had real CPU problems too (i5 4690k), especially after one of the more recent updates. What happened in my case was that the CPU usage for all 4 cores was flat lining at 100% with Odyssey running leaving nothing for other background processes that I needed like the joystick curves and head tracking software. What I had to do was specify CPU affinity (using a utility called Process Lasso), telling Odyssey to use CPU's 0-2 and my other bits of software to use CPU 3. Anyway, that kinda worked (barely) but eventually I was fortunate enough to be able to get myself a new PC.

Not sure any of this helps you much but you could at least try setting CPU affinity to tell Odyssey to stay off some your cores (which could at least theoretically help)?

The other reason I chipped in here was to say that while I was going through all this I learned that turning your graphics settings down can actually add to the load on the CPU (because the GPU is no longer the bottleneck) so there is a chance that turning graphics settings up might actually make your CPU less likely to overheat.

This is a really helpful answer, thank you. I am playing Horizons by the way, and definitely WOULD be fine with playing an older version of it. I don't know how to rollback though. Any idea?

I'll look into telling the game to only use cores 0-3 or something. I have eight cores, they may as well be used!

And yes, I'll put the graphics settings back to what they were. That's a great idea, thank you.
 
This is a really helpful answer, thank you. I am playing Horizons by the way, and definitely WOULD be fine with playing an older version of it. I don't know how to rollback though. Any idea?
I forget where it is but somewhere in the launcher is the option to pick the Legacy client (which you'll need to install separately to your current live Horizons 4.0 client) and play that instead. Note that your commander will only have the assets, rank and location from when the two game clients were split (some time ago now) and also note that legacy really is in maintenance mode (so you won't get new CG's, Galnets, ships, updates, etc), it's essentially frozen (although in a pretty solid and stable condition and with its own fully functioning BGS).
 
Last edited:
My computer's pretty old. FX-8350 [stock cooler], 16GB RAM, RX 570. Runs Ubuntu. I've been playing ED for a few months, 200 hours or so. Not many. I noticed relatively early on that I was getting thermal shutdowns occasionally, esp around starports. I put this down to the CPU overheating because of the stock cooler rather than the GPU or the 550W PSU. I also think it's the CPU because the PC shutdown the other day just when processing the Vulkan shaders on Steam before launching the game!

You should probably check temperatures during gamplay with something like HWiNFO.

Anyway, the FX-8350 is quite a power hungry part, most AM3 motherboards are rather weak when it comes to their power delivery, and an unnamed 550w PSU from a decade ago doesn't bode well either.

If temperatures are elevated, cleaning the system thoroughly would be the first step. You can also reduce clock speeds and/or voltages to get a bit more useful life out of a failing system.

I have put all graphics setting to low, to no avail.

If you aren't capping frame rate, reducing graphics settings may not be reducing load much.

Since I can't get a new system, what does anything think the LEAST CPU-intensive parts of gameplay are? Being in and around starports is a no-no. But I've noticed that in far-out systems everything is fine.

Is there anything I should do or anywhere I can go where I can hang out for a longer amount of time without the thing shutting down?

Any place that gets higher frame rates uncapped is doing less work per frame and will be less stressful if manually frame rate limited.
 
I'm no computer expert but if my pc were shutting down, i would check the temps of my hardware using CPUID HWMonitor. If it shows that the cpu is overheating i would apply new thermal paste and clean the pc of dust (if any) including the cpu heat sink whilst it is removed from the motherboard to apply the paste.
Then i would set my monitor to 60hz refresh, cap frames at 0.01 less than the monitors actual refresh rate using rtss (there are sites which measure a monitors actual refresh rate). This may give a smoother frame time.
In game 60hz, vsync on and play around with graphic settings so as to maintain 60 frames per second.
This is just my opinion as i am no expert and it could be another component shutting down your pc.
Good luck and i very much hope you get it sorted. o7
 
an unnamed 550w PSU from a decade ago doesn't bode well either.
Assuming temps are alright, PSU is most likely the culprit of random shutdowns under load. I've had one go bad, replacing it fixed random shutdowns for me. In fact, I'd replace any decade old PSU in principle to avoid the risk of it going boom and taking out the rest of my PC components. Considering GPU and sometimes CPU prices, better safe than sorry is my motto lol
 
In b4 "get a new computer" which absolutely IS the right answer. I can't afford a new one though [sadz].

My computer's pretty old. FX-8350 [stock cooler], 16GB RAM, RX 570. Runs Ubuntu. I've been playing ED for a few months, 200 hours or so. Not many. I noticed relatively early on that I was getting thermal shutdowns occasionally, esp around starports. I put this down to the CPU overheating because of the stock cooler rather than the GPU or the 550W PSU. I also think it's the CPU because the PC shutdown the other day just when processing the Vulkan shaders on Steam before launching the game!
1. Remove dust from inside computer. Use bicycle pump for example.
2. I had "shutdown problem" on 3 different LAPTOPS using Linux. It turns out, they have special mode "cooler boost" (with a lot of the noise) and Linux does not turn it on. So I would recommend to pay attention on cooler's speed and research this topic.
3. On desktop you can replace coolers or add machine oil using syringe into it. Try to update heat sink under CPU (but it needs some skill, for example, I don't risk to do this on laptops).

Nowadays, just booted computer (air cooling) should be around 44C. If it has dust it will be over 55C - time to blow it.
 
Alright, I didn't see anyone suggest this so let me tell you about a problem I had for about a year. My PC was working fine, and I did a hardware update but it was concurrent with a software update Nvidia did and some software installation I did. So comp would work fine, but when I started a 3D type game, it would work for a few seconds... The. The comp would force restart.

I thought it may be a graphics issue, I ran stress test and it would fail. The more I ran I began to notice it was at a certain temperature. So because of this I thought it was a graphics card issue. Changed graphics card. Problem persisted. Someone suggested PSU was going bad and was not able to take the heavy load when 3D game started... Changed that. Problem persisted... Bought new CPU and fan... Didn't install it. On a whim I re thought about the events that occured when the problems began. I had changed RAM a couple of months before the problems actually began. I had bought two different brand sticks of RAM that were supposed to be compatible with the ones I had. I took them out. Put on the games ... It has been working fine since then.
So check your RAM. Might be loose, or if it is recently upgraded it might be your problem especially if mismatch.
 

rootsrat

Volunteer Moderator
Another option would be to cap the FPS to say 30 and then manually edit the configuration files for Low preset to set the values even lower.
Not sure how playable the game would be then...
 
I would recommend always running a temperature monitoring software, such as MSI Afterburner, which shows relevant temperatures on the corner of the screen even in-game.

As long as CPU temperatures stay below about 85°C and GPU temperatures below about 80°C, it should be fine. (Of course the lower the better, but even at those temperatures it will still be fine.)

If CPU temperatures reach or surpass 95°C, or GPU temperatures reach (and especially if they surpass) 85°C, you should start worrying.
 
I would recommend always running a temperature monitoring software, such as MSI Afterburner, which shows relevant temperatures on the corner of the screen even in-game.

As long as CPU temperatures stay below about 85°C and GPU temperatures below about 80°C, it should be fine. (Of course the lower the better, but even at those temperatures it will still be fine.)

If CPU temperatures reach or surpass 95°C, or GPU temperatures reach (and especially if they surpass) 85°C, you should start worrying.

Thank you, this is exactly what I did, by running watch -n 1 sensors in Linux. At the moment you can see the idle CPU temp [k10temp] is very low, and never goes above 36C when playing Elite these days as I have the new cooler.

1747480197374.png

With the old cooler, it would go above 80C during play, and then the PC would shut down.
 
We still have an FX-8350 on the network

DNS Name

"Flamethrower"

Keeps the basement nice and warm at idle

Hah! Yes, it does get pretty toasty with a stock CPU.

As an aside, I checked out the book list in your link and was pleased to see I'd read both the first two - 1984 and Brave New World. Both excellent.

PXL_20250517_142542976.MP4.jpg


I find my 1973 copy of BNW to be suitably hipster [sorry for hijacking my own topic]!
 
Back
Top Bottom