How about analyzing the frame rates of both clients during the hot run and the cooler run with the clients FPS monitor?
My hunch is that V-Sync does not get respected at the first run and the GPU will go on a frame calculation spree until it hits throttle. V-Sync is set within EVE and also within the GPU driver settings and both are also dependent on the displays FPS settings.
Thing is… the worst this would do is load up the card to max, at which point it should still be more like 65-70c, not redlined.
That it is even capable of hitting 83c definitively tells you that there is a problem dissipating heat.
Hence me going “check fan speeds”.
If this is only happening to you and no one else, I’m having trouble understanding why you want to submit a ticket asking CCP to change their software for everyone just because your computer is malfunctioning. It’s not Eve Online. It’s your computer whether that’s hardware, firmware, or your OS.
And no one is going to believe it’s not your hardware unless you start posting some specs like people have asked.
1 Like
You’ve obviously never run a Universe Sandbox multi-moon collision.
My goal would be at first to be getting physical values during the actual behavior for a discussion of direction of the possible cause instead of getting readings from a crystal ball… even if the GPU is not working as intended, you could still keep the hardware unchanged if you knew what happens and implement a workaround for it.
Well I gave you the way to test that as one of the options up there. I didn’t leave it out, I do a lot of game testing, and have found some do mess with your hardware. They do this unintentionally by using code, compiled on their system, and not yours. A lot of the developers don’t seem to care much about backwards compatibility or compatibility with any other machine than the ones they own.
I test on machines as far back as 2015. Some businesses will have old hardware laying around a very long time. When I contracted for GM, they were still using Windows 3.11 and this was back in 2001. I suggested they upgrade to XP as it was the new kid on the block. As my contract finished that summer they decided to finally upgrade to Windows 98. Real genius at work making cars.
Try out what I told you, let me know if it makes any difference.
Newer CPUs/GPUs can handle 100ºC. Even an Intel i7 from 9 years back is okay up to 100ºC. 83ºC is only a problem on old chips.
That is a typical value given for tjmax, which is basically the point at which you’re in danger of damaging it.
83-85 is where every modern Nvidia card begins to throttle itself (again, if stock; this is all adjustable but I don’t get the impression OP is doing anything under the hood), including the very latest models. It is not in danger of damage at that temp but it is reducing performance to keep from going above that temp.
You haven’t answered whether the issue happens if you only run one toon. This is a rather annoying thread in which you’ve actually given very little info on your setup and what you are actually doing.
Then you’re wrong because it actually did overheat a lot of players computers, in fact CCP had to optimize that code a few times just to resolve the problem… Also it seems you’ve forgotten all about the ‘Boot.ini’ fiasco…
There’s other examples too but yeah, let’s just continue saying CCP’s programmers and the work they do is excellent and the problem is actually everybody else’s computer… 
I appreciate that you believe that, but you believe a lot of silly things.
Also it seems you’ve forgotten all about the ‘Boot.ini’ fiasco…
Boot.ini was a real thing that is also wholly divorced from magically making a GPU overheat.
It’s not a matter of coding excellence. There just… isn’t really anything you can do from client software to overheat a GPU (again, even the New World Vs. 3090s was determined to be a fault in the GPUs).
If you believe otherwise, please share some code capable of overheating an otherwise functional GPU through directx draw calls.
I will wait.
yeah, sorry, just haven’t been back on the forum since - and haven’t tried just one toon - will try now
I didn’t mention the setup to avoid confusing the issue, since the system DOES run cool except when I first bring EVE up after re-boot.
I should mention that if I minimize either of the screens, the temp falls to about 74 in the high-heat scenario.
**** update
well, I did try running just one toon - but some weird stuff happening and here’s some more input.
I normally run my boxes 24/7 - the scenario I’ve been describing has been the case since I noticed the high temp readings a week or so ago - and when I restart both toons (not the computer) it clears up and runs cool.
BUT not today - when I restarted the toons, the 83 repeated - isn’t that the way it goes LOL?
Anyway, I took everything down and this time I re-booted - AND surprise, both toons started up and the box is running 60. So, all my testing data so far on the issue OUT THE WINDOW.
I appreciate the attempted help - I’ll just have to do some extensive testing with the info y’all have given me so far and get back - reporting these issues piecemeal is just going to waste y’alls’ time but thnx again!
Obviously you weren’t around during that time so take your silly talk somewhere else…
Seriously, can you not understand that was just an example to show what bad programming can do?
Anyway, I don’t really care what you have to say nor do I want to interact with you, especially since you immediately started your reply with sarcastic demeaning remarks…
I have been around longer than you have.
Bad programming can delete a file, but it can’t defy the laws of thermodynamics, so magically generating heat in a GPU isn’t within the realm of what can be done via bad programming.
Ok Mr Wizard…
AI Overview
EVE Online can cause GPU overheating due to high frame rates (FPS) in stations and intensive 3D rendering, a known issue often linked to game patches. Key fixes include enabling V-Sync to cap frames, limiting FPS in settings, updating drivers, and cleaning system fans. Other solutions involve reducing graphics settings or using
CTRL+SHIFT+F9to disable 3D elements.Key Causes and Solutions
- Station Overheating: The EVE Online forums show that being docked in stations can make GPUs run hotter than other AAA games due to excessive FPS.
- V-Sync Fix: Enabling V-Sync (Interval 1) limits FPS to your monitor’s refresh rate, significantly lowering GPU usage and reducing temperatures.
- Limit Frame Rate: Using the in-game “Advanced” settings to cap frame rates prevents the GPU from working unnecessarily hard.
- Lower Graphics Settings: Reducing graphics from “High/Ultra” reduces the rendering load on the GPU.
- Disable 3D Elements: Pressing
CTRL+SHIFT+F9disables 3D graphics, which can immediately lower GPU load.- System Maintenance: Cleaning dust from PC vents and improving airflow is crucial to prevent overheating.
Note: GPU temperatures between 65°C and 85°C are common during gaming, but exceeding 90°C is generally considered dangerous.
GPU overheating are common among players, often attributed to the game’s high frame rates in specific environments—like stations and hangars—where limited graphical complexity allows the GPU to render hundreds of frames per second, leading to 100% utilization.
Common Causes
- Uncapped Frame Rates: If the in-game “Interval” setting is not configured correctly, the game can run at unlimited FPS, maximizing GPU load even in visually simple scenes.
- Station Rendering: The “ship spinning” view in stations is notoriously intensive, with some players reporting higher temperatures while docked than while in space.
- Graphical Updates: Recent shifts toward a “GPU-driven pipeline” and the introduction of volumetric effects (e.g., in the Equinox update) have increased the demand on graphics hardware.
- DirectX Version Issues: Some users report higher GPU usage with DirectX 11, though results from switching to DirectX 12 vary by hardware.
This is AI slop making circular references to a forum where people who don’t know better think a game can overheat their GPU.
For all practical purposes, you may as well be citing your own incorrect posts on the subject.
Doesn’t matter if it’s AI generated, you’re wrong in stating that GPU’s can’t overheat due to game code…
Anyway, good luck in your programming career… 
My old GTX 295 (single card SLI from years ago) operated at 94C, even after me trying to under volt and under clock. This got me worried so after some internet searches the conclusion was that it was fairly normal and was fine (single card SLI was a really bad idea) so I tested it to see at what point it would die on me, turns out that 104C was the upper limit where it would conk out.
Given Moore’s law and my “need” to upgrade often I decided I wasn’t going to worry about longevity, I clocked it and gave it the beans. Ran fine at 97C until it got replaced.
It can’t, though, and your post doesn’t show otherwise, though given your critical thinking skills, it’s little surprise that you imagine it does.