Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Discussions Rules and Guidelines
Report this post
I have 2070 super and it runs round about 70 playing this in this weather. I also have a 240mm AIO cooled 9600KF which gets to just over 60 during same period.
I did have cooling issues until I realised I had mounted radiator fans the wrong way around.
If you discount overheating you will be getting closer to solving things.
I play this game at 2k, max graphic settings with no problems.
I confirm my amd processor rises to temperatures beyond its habits, even Cyberpunk for example in the middle of town with all the options on high does not do this to me.
Dark Descent is more demanding than the AAA triples I've played, it lacks 6 months of graphic development.
Monitoring the behaviour with HWinfo, it's only when browsing the menus extensively while in a mission that temperatures go up (even with capped framerate). Framerate even goes down in them.
I've also seen the lighting bugging/changing when browsing through the menus and then going to the marine status screen where it's the 2D/3D mix.
Closing and opening the map usually fixes it again.
Something is busted with the menus when in missions imho and I personally use the map alot which is usually fine. But as soon as I started browsing through the codex or datapads, something's going awry.
Nothing is overheating or crashing though on my end. Only crashes I ever had were due to mechanics*
(* putting mines under motion trackers and firing up the tracker causes instant crashes for me, so I reported it and stopped doing that lol).
Temps are reported in Celsius, even for us Americans, so I'll be using that below. Offhand those temps seem fine to me. Capping FPS is really just going to reduce the load on the GPU. It's never the correct way to "fix" thermal issues and really just either a workaround or a way of saying "I don't need higher FPS than X". So if capping FPS helps I'd immediately suspect thermal or power issues. Systems are only considered stable if they can consistently run under full load (100% usage of both CPU and GPU). If they cannot there's a problem. Reducing usage is a workaround ignoring the real problem.
It depends on the CPU and GPU specs, and modern CPUs have even pushed this limit, but typically thermal shutdown happens around 100 C and a bit north. Usually I'd say aim for under 90 C on both, but there are some modern CPUs that try to peg themselves at 95 C. In most cases though it's fair to guess if you hit 100 you've got problems. If stability isn't already an issue. When working on someone else's hardware if I see numbers hit 90 under load I immediately presume there's an issue (or one not far off). Lower is virtually always better.
When temps get high two things happen typically, in this order. First there's thermal throttling. Temperature is basically inefficient use of power. It's transfer of energy where some is lost and converted to heat, much like a combustion engine in a car. So heat is bad, both because it ruins things and because it literally means we're wasting energy.
Signs of thermal throttling are slower clock speeds. In effect a CPU or GPU will try to generate less heat, but lose performance, to maintain stability. Thermal throttling is never a good thing. In that case you have expensive hardware that is losing performance only because of heat.
If you see thermal throttling, or high temps, I wouldn't freak out though. This is actually far easier to resolve than most would think in most cases, unless you're trying to overclock like myself in which case you've signed up for the responsibility of "I know what I'm doing and accept risks".
Air duster is your best friend with computers, especially laptops which lack efficient cooling. After that you have things like better airflow (such as some new case fans, which are remarkably cheap). You'd be amazed how much difference adding a single new intake or exhaust fan can help airflow, and the cost is typically very cheap.
In the worst case the cooling solution is inadequate. For a CPU that's typically swapping out for a better air cooler or upgrading to water cooling. If you're brand new to water cooling stick with an AIO (all in one). Those work out of the box without having to build your own loop which is much more daunting.
Do note, as others have mentioned, oftentimes thermal adhesive (usually referred to as paste but not always the case anymore with things like liquid silver and thermal pads) is incorrectly applied or simply insufficient. This is FAR more common than it should be. So if you didn't do it yourself, even if it's not pressing, I'd put it on the list of things to do.
If you have a pre-built and did not apply the thermal adhesive yourself this is probably a good experience both to learn and to get the satisfaction. You can buy very good thermal paste for a few dollars. I will redirect to a good reference, but the biggest mistake one could make is not removing any existing compound. So if you're upgrading (or fixing) the thermal adhesive, usually it's new paste, grab some iospropyl alcohol and that'll clean it up real quick. It is never, ever, safe to apply new thermal paste if there's some existing. Even if it's the same. You always need to remove what's there first.
If thermal throttling fails, then we hit thermal shutdown. Presumably before stability fails, but not always. Crashes, BSOD, visual artifacts and other things are all apt to happen if thermals get far too high. The good news is the BIOS (firmware on the motherboard) is ultimately responsible for all hardware. So it does its job well and pulls all power if it detects hardware over thermal limits. This is know as a hard shutdown, it certainly doesn't wait Windows or whatever OS to do it's normal shutdown sequence.
The good news is if you see high temps that hit thermal throttling or thermal shutdown the BIOS typically does a good job from letting hardware fail. So you can generally speaking sleep soundly knowing nothing is going to fry. Computers are not cheap.
The bad news is if you see either it really needs to be addressed. I'd always keep air duster handy and use that. It's not even optional if you own a laptop. After that it's simple things like making sure all fans are running as expected (and are correctly doing intake or exhaust).
Probably the best news of all is that addressing a thermal issue is often very cheap. We're talking under $10 for new thermal paste. Under $20 for a new fan. And also not much more for brand new air cooler, maybe a tad higher if you have a power hungry CPU. Water cooling is getting close to the standard for high end CPUs and almost a necessity for overclocking and outperforms air cooling. But a bit more expensive and not as simple as air.
It can just be a little scary or hard to understand for beginners, but there is ample knowledge to help people out. Oftentimes I'm happy to find the issue is thermal related, as diagnosing software and drivers and firmware is usually more time consuming. The absolute worst scenario with thermal issues is you'll need a new case, but that's rarely the problem unless you built it yourself or the prebuilt was designed horribly.
Let me stress one thing. GPUs and CPUs run very, very hot. Some are even designed to hit a certain temp. It's very common to see a GPU operate between 80 and 90 C all the time. Same goes for CPUs, but I try to aim lower than that. When in doubt, first look up the specs of your CPU and GPU and safe thermals should be easy to come by. As well as hard thermal limits (shutdown). So in short, never look at a number and go "oh no".
The BIOS is there to make sure you aren't damaging hardware. And improving thermals is a fun and relatively easy way to learn how to maintain and build your own PC. Nothing is more rewarding than doing some work and then booting up, putting the CPU/GPU under load and seeing lower temps. It's far more often that laptops suffer heat issues (especially over time) but if you've ever built a new PC or bought one this is not uncommon at all. It's one of those things you address every so often then it's smooth sailing.
Another common cause on the hardware front is the PSU (power supply unit). These can go bad or just fail to provide adequate power. I literally just ran into this issue myself when upgrading GPUs to a 7900 XTX. It ran fine under load at pretty whopping 300 to 400 watts, but transient spikes would cause hard shutdowns. These are very, very brief spikes in power demand that even a perfectly good PSU might just fail to meet. And then there's just PSU failure. This is far more rare, but still notably common enough to mention it.
Once you can eliminate heat and power as issues, almost all problems are usually traced to software. But I must stress, like I do every time I comment on this, software cannot cause thermal issues. I, as a coder, cannot force your CPU to fry. That's under the control of the BIOS really. So anyone that blames a piece of software for thermal issues is incorrect. Even IF that software causes thermal issues it's only unmasking an existing problem. I run an overclocked CPU, GPU and RAM. Wayyyyyy past its limits. If software alone could cause heat overclockers would be screwed. Much less everyone else.
A great place to start if you want to learn more about your CPU, GPU, motherboard, thermals, case, etc... is Gamer's Nexus. They have PHENOMINAL unbiased coverage of all of this. They can help you understand various things and make wise purchasing decisions so it's a great start from absolute beginners to very advanced. You can find their channel here, And no I'm not affiliated with them. Just extremely reputable in this space, there's a ton of resources out there to get other opinions. Hope this helped somewhat.
Finally, and I cannot emphasize this enough, anyone who tells you software like a game can cause overheating has NO idea what they are talking about and are very, very wrong. Don't listen to their advice and immediately run screaming. Unless you hate yourself like me. Then try to correct them and fail miserably. Or simply ignore it.
https://www.youtube.com/@GamersNexus
Careful doing this. Not all PCI slots are the same on a motherboard. Typically the top slot is the fastest. That's the real reason you'll almost always see a GPU in that slot. It's designed for it. This does vary between boards though, so if in doubt consult a manual for your specific motherboard.
LOL. It's not April 1st. Made my day though.
Your computer should be able to run the CPU and GPU at 100%, all day long. All you can do is make sure there is no dust increasing temps and add adequate cooling to the component causing the crash (if indeed your temps are causing a hard crash). If you cannot do this, I think the only option is to reduce graphics settings and\or cap the frame rate to something less. [edit: Spelling]
I have a 3070 Ti and a Ryzen 9 5900X with 32 GB RAM. I play maxed out at native 2560 x 1600 (no FSR garbage). Locked 60 fps. No stability issues. Temps are good. GPU Utilization is around 80% and CPU utilization on the 5900X changes between 15-30% depending on how I zoom the map etc.
I have in-game vsync off and vsync enabled in the Nvidia control panel.
Edit: All components in your PC should be able to run at 100% utilization without crashing. If they don't, your PC isn't stable.
I can run my CPU and GPU for hours and hours under full load, doing 3D rendering. That's what a PC is supposed to be able to handle.
Hey, nice info you add up there Zero. Useful for the inexperienced or badly misguided user to learn.
Sadly I doubt a stray rare soul would finish or comprehend all of that. For a beginner level of course.
I'm gonna add that. There are also other specific problem that capping the fps almost fix all redundant problem too.
- Some game engine by designed level tied a script per frame such as Fallout. Having too low fps can cause npc stop moving in tense area. Too many things happen causes a delay ed script.
- Some game having too high break a script which cause strange behavior especially on older engine. NPC run inhuman fast, cut scene movie perform ridiculous behavior, door is not opening (Poor Old Dead Space), game get stuck at certain place. Hench force fps cap fix it.
- Worst part is some sync technology monitor doesn't play well with some engine for early adopter buyer. Game force crash, strange fps running too low and swing than normal. etc. You can dig this info in both AMD and NVIDIA official forum. Been a years. Avoids using it like a plague during that time.
------
For what it worth, playing old game or in some situation that the game forget to put a cap for you (Especially during cut scene, in-game menus, etc). Makes you reaching over 500+ or near thousand frame doesn't gonna make it smart stress and healthy to your ears. For typical user of course. Capping it is your best bet.
Thanks. I do agree with this as well. Regarding capping FPS I only meant that it's never the right call to do for temps alone. That's masking an issue. But there are many legitimate reasons to do so and you gave great examples.