GPU: Radeon RX480 8GB
CPU: AMD FX8350 4GHz
Motherboard: Gigabyte 970a-DS3P
Memory Clock: 1266MHz
OS: Windows 10 Home 64-bit
Graphics Driver: Crimson ReLive 16.12.1
No overclocking, it's all 'Out of the Box'
So, firstly I'm hoping for some sensible and helpful discussion. I've tried several 'gaming' forums and just received abuse about clearly how stupid I am...
Anyway, I've bought a new PC and been having an issue with a few games with crashing. The games I've played so far:
Dishonored 1 (about 30 hours) on 'Ultra' settings - No Issues
Rome 2 Total War (About 15 hours) on 'Extreme' - No issues
Elder Scrolls Online on all settings - crashes
DOOM on all settings - crashes
The crash itself regularly occurs about 10 minutes into the game, at which point the screen goes blank, the audio goes dead and effectively my PC locks up. The only way to recover from this state is to Hard Reset the PC. At which point it is running OK, with NO crash logs or error reports suggesting what the problem was.
So I've tried all the usual troubleshooting methods (re-installs, update all drivers, turn off firewalls etc.) nothing makes any differences. 10 mins in and CRASH. So I've been monitoring my hardware in game. All games run perfectly smoothly, well over 60FPS and stable; no stutters or drops even on the highest settings. I did notice the temperature of the GPU getting to around 80°C. This is where the abuse came in; many people have told me that 80°C is perfectly normal, acceptable, no issue and my problem is elsewhere, but I decided to follow up on it anyway. Low and behold if I manually juice up the fan speed in the RX480 so it maintains a steady 55-60°C no more crashes and no more issue. In my mind it could make sense, as the games that are crashing are newer and have better graphics so are no doubt putting greater strain on the GPU.
So I have a few questions for the AMD community (where better to ask!?)
1/ IS 80°C OK, or is it really getting a bit warm?
2/ Is this sort of crash indicative of a heat related issue, or could I have another problem?
3/ If I can turn the fault off an on again with fan speed; why is it necessary to use 'Manual' controls? Are the automatic fan settings not up to scratch?
Would really appreciate peoples thoughts and comments. It's a brand new PC, so I'm wrestling at the moment whether to return it (though I'd rather not), or whether I'm just being a bit amateur (new to this PC gaming business!)
Make/model of card(or is it a reference single fan card) ?
Case mini/mid/full size tower, how many case fans ?
There's this for Doom tweaktown.com/guides/7757/doom-graphics-performance-tweak-guide/index.html
Hello, Just like to chime in here, you didn't mention your PSU, and/or what you might be using for a power scheme on your desktop. And that's a pretty hefty card, what is your monitor that you are using and the Resolution?
It's the bog standard make I believe. Single fan.
The case is mid size, its a Neptune ATX gaming case, if that helps. Has only one fan at the front. However, I am running an aftermarket cooler on the CPU (Cooler Master Hyper 103) which is excellent at keeping my CPU cool. It rarely gets above 50°C.
With regards to the graphics tweaking, I see what you're getting at, and I did some experimenting. Interestingly enough there's no a massive end impact on the heat whether the settings are Low or Ultra. If I leave the fan control on auto, it quickly gets up to 80°C and crashes.
Hi Ambrose. Sorry, a glaring omission from my list. It's a EVGA 600W power supply. I know the card is a bit of a beast. I did a few calcs (Fans, CPU, GPU etc.) and I don't think I'll be exceeding the supply. I have been monitoring Power in-game and at max load (when loads of stuff is going on) it's just tips over 100W. Think that could be an issue? If it was supply related would the PC just hang, or reset altogether?
I have selected the High Performance scheme
My monitor is a 24" BENQ (RL2455HM) and I'm running it at 1080p 70Hz refresh rate.
I don't use W10 so I'll just ask some questions and maybe it might help you come to a reasonable solution path for this. You mention it only has 1 fan at the front, how do you feel about the internals of your tower and it's ventilation? the fan on the cooler , the fan on your PSU when it's all going ( When it needs to ) do you think the front fan is pushing enough of the internal air to vent out of your tower? Have you tried running games with the case open? Has or does it make a difference?
Again just throwing ideas out here.
You may be better moving (or starting a thread) in the Graphics section, will get more views/replies ?
I don't have a 400 series or amd setup.
There's a couple of possible issues with your build (imo), was it bought prebuilt, or you chose components ?
Mobo drivers/bios up to date(there are two revisions of that board (one old, one with recent updates)).
Which also leads me to query what exact model PSU (a couple of older ones were barely passable in their day).
Unless your living in year round cool climate, 1 (exhaust ?) fan is insufficient for gaming.
For the ref. model card, you might want to check the links in Tip to improve RX 480 performance: UNDERVOLT IT! may help.
Were there problems before the new driver/UI software ?.
I'll have a go at answering your points, but I am very new to PC ownership so finding it difficult to gauge. From my point of view the level of ventilation is OK, it has a side panel with holes, so plenty of air gets in. The CPU cooler is excellent (Cooler Master Hyper 103), and keeps the CPU plenty cool enough (rarely gets above 50°C). Actually the in-built case fan seems a little pointless me, as it's at the front of the case, where all the warmest stuff is at the back. It doesn't really seem to be kicking out warm air when everything is running.
But in terms of the whole heat management thing, I can prevent crashes, and keep the GPU temperature below 60°C in game without issue; I've just had to create my own, more aggressive, fan strategy for the GPU itself. This is what seems odd to me 1) Why is the automatic control not fit for purpose? 2) Why such a lock-up/crash/freeze due to excess heat? is that normal? why doesn't it just crash the offending program?
Hello Goodplay. Good suggestion, I'll do that too!
In terms of the build, I specced the components (after doing extensive research on compatibility), so everything is compatible from a motherboard-CPU-GPU point of view. However I did not build it myself, I got it assembled and tested by a PC specialist company. I reasoned that they would be less likely to make a mistake, would have more experience at doing performance/stress testing & I would get a warranty on the build.
Not 100% sure how to check the bios is up to date? All the hardware drivers are as I have checked via Windows Device Manager, but how do you check the bios.
I'm coming to the conclusion I agree entirely about the state of the internal cooling, and have this morning ordered some new fans. Hopefully they can make a difference.
The problems have been since Day 1, but I've only had the PC for 1 month. It arrived as a fresh install with all drivers up to date.
I'll keep this short, You mention that it is a new computer. Is it possible to get any help via where you bought your computer? I am surprised that they wouldn't have tried to inform you that you might need better ventilation ( More fans, or different Tower ) due to the Video card you chose to install. Or was the new card an upgrade to what they might have had from a default purchase?
Cheers, Please keep us informed how it is going.