Hi, I just got the new RX 590 nitro+ SE but I'm already having issues and I was wondering whether or not I should return it. I've been experiencing graphics driver crash after a few minutes of playing games like BF1 or BO4, so I've tested multiple stuff to try and fix it, and I believe using a custom fan speed to not go over 50°C resolved the issue. Keep in mind that this card in idle with two monitors is at 50° C. But that means my card has a problem with high temperatures, when I use the automatic fan mode the temperatures reach 70° C + , which is normal for a gpu. Also, in BF1, where I'm supposed to get 115 fps avg, I can barely keep up a steady 75fps when I cap it for my freesync. So yeah, should I return it ? EDIT: I also tried with Witcher 3 at max settings with or without custom fan mode, same things happen. Here's an image of the trixx hardware monitor while using custom fan mode. https://imgur.com/a/QJfQQCu
No one needs two monitors in the first place. Begin from there. If cleaning up gpu driver (with DDU) etc and removing obsolete monitor from device manager fixes it your are OK. If not, RMA, or go Nvidia, your gpu is not supported in older AMD drivers to test unfortunately.
I did all those things already and none worked sadly. I guess I'll wait a few days, incase AMD releases new drivers and it ends up being a driver issue and not a hardware issue.
That's the best for now. Also don't forget more than one monitor gives different idle clocks, idle temps and maximum Temps during game play. If it keeps being a problem you may even consider different aib partner than shappire next time if you return it. Also check for newer vga bios or contact sapphire support. .
It is very unlikely to be driver issue. Unless driver improperly controls voltage. Is wattman enabled? If no, then card voltage should be in default mode as written in vBIOS. Driver crashes are more often than not caused by HW instability. There is still good chance that Card is OK, but your PSU is not good enough. DO you have that 520W Antec which is in your sig? Then it may explain reduced performance too.
Temps are fine if they're at 70°c under load is a normal temperature really, especially if a pc has poor airflow (they are slightly higher than you would expect but not that bad, I'd be more worried if they were 80°c and above) as for idle temps, 50°c isn't really a problem, modern cards tend to idle at higher temperatures these days because of things like zero fan technology (i.e. they turn the fans off until the card reaches a certain temperature which can be as high as 60°c). The best thing to do is just test everything under defaults, remove any overclocks on your system. It might also seem odd but sometimes monitoring software can cause problems too, especially with new devices so try without using trixx or afterburner for the purposes of stability. If you're still having problems it could also be another issue, perhaps PSU related? Checking your profile, you have a 520W Antec PSU? Would be ideal for you to give out complete specs of your pc. If the card is at fault though, definitely do exchange it, perhaps you're unlucky to get a faulty card (quite unusual for a Nitro+ model to be faulty though). Depending on where you purchased your card, if at a retailer, you could easily get it exchanged, if online, you're probably going to have to go the RMA route - which country are you from? If in the UK (and other European countries), you can just get an exchange or refund where you bought it from without needing to go through manufacturer RMA procedure.
Just run a stress-test Try a 1,2 hour pass of AIDA's GPU-test and if it fails - open an RMA with your findings plus your attempted AIDA pass. Your power supply is fine provided it supplies 30 amps along the +12V rail(s) (and of-course is in good-working non-defective order)
Thanks for all the suggestions, I'm going to test it again with my RAM and CPU OC disabled, also I have 2 SSD, 1 HDD, and the specs from my profile, and since the recommended PSU for this GPU is 500W+, I thought my 520W Antec was fine. I'll keep you updated. EDIT: still not working without my OCs, also about the PSU, if it was an issue with it, why do I have no issue when I use my custom fan mode ?
This is a whole system power consumption with a power hungry 8700k at 5ghz. Your PSU is fine but seems like card is acting up. https://i.**********/cJNZ1r6J/Untitled.png
Could be misaligned heat sink placement causing the memory to overheat. You could try downloading hwinfo64 and running the sensors only tab. Look at the GPU memory errors section and the GPU VRM voltage and temps sections. It can also be easily linked to msi afterburner overlay allowing you to monitor said settings in game. Simply run afterburner first. HWinfo64 https://www.guru3d.com/files-details/hwinfo64-download.html MSI Afterburner https://www.guru3d.com/files-details/msi-afterburner-beta-download.html Edit. Also keep in mind, not clearing out any Nvidia driver remnants can cause issues, although this does appear to be a temp related problem.
RX590 is not worth it . I have the Powercolor rx580 Golden Sample overclocked to 1425mhz and I bought a 2nd one and run them in crossfire and I can play anything on high or extreme and have no lag .
If you have a friend's system to test the card in you could try that as a means to isolate everything else - bring your card try it in that system if it persists, there you go - otherwise tests like AIDA stress the card for errors and offer trials you can load in a WindowsPE environment via a USB stick if you don't own a license.
Yes I cleaned everything with DDU multiples times already, do you know why I can't see the GPU VRM temps section in my HWiNFO ? I only see the voltage
They might not have updated it for the RX 590 yet, but you could try GPU-Z. GPU-Z https://www.techpowerup.com/gpuz/ I would also suggest putting your system clocks to default, if you haven't already, and in particular paying close attention to your system memory. A new graphics card will always put more stress on your components and may show any weaknesses with overclocks etc.
I tried with or without my CPU and RAM OCs and I had with both the crash issue. Also here are the logs of HWiNFO and GPU-Z while playing The Witcher 3 until it crashes (in the middle of the logs you see the gpu load going to 0% it's because I got a pop-up and it brought me back to the desktop).
There's no logs showing? Edit. Anyway, I would also clean the wafer part and re-seat the GPU, make sure all the pci-e power cables are properly connected, and if using two cables, make sure each is coming off of a separate 12v PSU rail. Not sure if your model of the trupower has a single rail or two rails, if the latter your documentation will show which connections are for separate rails. PSU age has a part to play too sometimes. Check the GPU-Z log to see if the VRM 12v rails were significantly dropping. It's a software reading so might not be 100% accurate but shouldn't be off by a mile. You can also checks your PSU's 12v rail in the Bios or by using hwinfo64.
