4 Replies Latest reply on Oct 8, 2015 11:58 PM by goodplay

    My potential fix for the 390 Driver Crash


      Hello everyone,


      First time posting here, so let me share my recent experience with the MSI 390.


      I have an i7 2600K with 8 GB DDR3 Ram, an Asus P8Z68-V Pro/Gen 3, an Evo 850 SSD and 4 HDDs, and my previous card was a faulty Asus GTX 970.


      Upon receiving the MSI 390, I installed it on this pc running Windows 7 64 bit after running DDU and CCleaner to get rid of previous drivers.

      My problems started shortly after, with random crashes on nearly every game and performance hiccups due to unstable FPS's, going anywhere from 300 to 20 in some games.

      Most crashes were solved by uninstalling the Gaming Evolved APP ( Raptor ), but recent titles such as Rainbow Six : Siege and COD: AW still crashed.


      I tried several different drivers with no success, before having to upgrade to Windows 10 and doing a fresh install of it on the new 850 Evo SSD.


      After updating Windows 10 and installing the latest 15.9.1 Beta driver, I noticed my FPS were much more stable, but I still got the driver crashes in the same games I previously had.

      I tried changing to the Stable driver, and the problems persisted, so I contacted a friend that has far more understanding of computer hardware than I do.


      He had me downloading the MSI Gaming App, which didn't work ( something about the platform not supporting this application ).

      He then had me downloading MSI Afterburner, in which we ran some tests and eventually "fixed" the problem.


      The default values were "Core Voltage : 0", "Power Limit : 0", "Core Clock : 1040", "Memory Clock : 1500".


      At first we lowered the Core Clock to 1000 MHz, which seemed to solve the issue if I ran the games with nothing else open, but if I had, for example, OBS recording they would eventually crash.


      After many tests, I ended up increasing the Core Voltage by small increments, noticing that it took longer for the driver to crash each time.

      I left the Core Voltage at +25 mV, and keeping the Core Clock at 1000 MHz, which has so far proven stable and I'm yet to get a driver crash under any circumstances.


      I now ask, could my card be defective?

      Could this be solved in the future by a vBios update?


      The computer shop I bought it from replied to my email, similar to this post, telling me to RMA the card, saying they had no previous complaints of the model, despite me seeing so many people with issues in these forums and throughout the internet.


      Did this solve the problem for you?


      Thanks for reading, I hope this topic helps.

        • Re: My potential fix for the 390 Driver Crash

          After playing around with my 7970 and noticing some things when playing around with the powerplay settings on it " TDP " things of that nature.


          I would like to note that I returned my Asus Strix OC CUIII 390X GPU after being told by Ray on here and other support specialist from here on this forum and or from Asus support that I, had some other problem going on and that it wasn't the card itself but something I'm doing or something failing in my setup..............


          I exchanged my card at the store for a new one per Asus request, installed new windows 10 and 7 installations over and over again reinstalled games and jumped through a thousand other hoops just to be told to try more things like switching out my motherboard for another one power supply etc. etc. etc.


          Which, is why I returned it to the store and got my money back. I personally think that after all this is said and done and it ends up being something like what I think it is at this point, I should get a free card from Asus for being put through what I've been put through, through out this whole ordeal.......


          Nothing but mental anguish and issue after issue and tons of wasted hours trouble shooting the issue for a couple of months to be told to trouble shoot even more.



          Sorry rant there,


          I personally think that the issues with the 390 series cards are stemming from some kind of bug relating to the TDP power limits or power play settings or power limit settings depending on how you decide to word it. I think that it's throttling itself to some extent or the TDP just isn't being handled correctly with applications for some reason, a bug perhaps.




          My 7970 is rocking hard right now without any issue at all.


          Here's a benchmark result that I had from the other night once I finally settled on a good 24/7 spot for clocks on my old 7970 card.


          AMD Radeon HD 7970 video card benchmark result - Intel Core i7-2700K Processor,ASUSTeK Computer INC. P8Z68 DELUXE/GEN3



          This result is from when I had the 390x in my system.


          AMD Radeon R9 290X video card benchmark result - Intel Core i7-2700K Processor,ASUSTeK Computer INC. P8Z68 DELUXE/GEN3



          Here's a side by side comparison of the two results.



          • Re: My potential fix for the 390 Driver Crash

            Have you posted this on MSI forum ?.

            Bios update would be supplied by manufacturer(MSI for yours), possible if enough users tried your possible fix/reported their findings.

            I would RMA card.

              • Re: My potential fix for the 390 Driver Crash

                I haven't due to spending more time doing troubleshooting, but will do so soon enough, thanks for the suggestion.


                To add to this post, I've come up with the same driver crash problem again, with the same games and with Star Wars Battlefront ( all DX11 titles, I believe ).


                I've tried different drivers, different voltages, lowering the clocks, running fullscreen and windowed, running in all settings, running while streaming to twitch, running with the browser closed, and so on.

                Every time I end up getting a driver crash, sometimes seconds after starting to play, other times 1 hour after.


                AMD is not handling this situation with transparency, in a timely manner, and is not respecting the costumers' time and money invested in troubleshooting their faulty products, so I will RMA the card and, if the problem isn't solved with the replacement, I'll simply return it and go back to Nvidia which, despite being more expensive, never gave me half of the headaches I'm getting with AMD.

                1 of 1 people found this helpful