The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
The Guiding Principles and New Rules document is now in effect.

Is it possible to have problems with one version of DirectX only?

I built a PC in January. I put in a rtx3060 12GB.
I keep the driver up to date.

7Daystodie was crashing sometimes when there were a high number of fire effects.

Cyberpunk worked but after the latest patch crashes on boot. I uninstalled it.

Dying Light 2 started crashing on boot after their latest patch.

Dying Light 2 was using Dx11, I forced it to Dx12 and haven't had a crash since.

7Daystodie was using dx11, I forced it to Dx10 and haven't had a crash yet but I don't think I have enough time in game to see if it has totally fixed it.

Is it possible to have a dx11 issue or am I seeing a pattern where there is none?

Posts

  • BahamutZEROBahamutZERO Registered User regular
    Mysterious set of symptoms. You didn't have a different GPU installed before this one, did you? Totally fresh build? My gut feeling is it's more likely to be a hardware problem than a directX problem. Try taking the GPU out and reinserting it in case it isn't mounted quite right, make sure there's no debris in the slot, that kind of thing. If problems persist, the card might be a lemon and need to be replaced under warranty.

    BahamutZERO.gif
  • wunderbarwunderbar What Have I Done? Registered User regular
    a more likely scenario is a driver issue. I'd completely remove the GPU drivers and install them from scratch again.

    XBL: thewunderbar PSN: thewunderbar NNID: thewunderbar Steam: wunderbar87 Twitter: wunderbar
  • RiboflavinRiboflavin Registered User regular
    edited January 30
    It was a new build. I wonder if I got a lemon video card. Built January 2023 I should have specified. I wonder what the warranty on those things are.

    The only thing from the old pc were the SSDs. I'll try re-seating the GPU, when I upgrade drivers is it not the same as removing them 1st? I just looked Cyberpunk requires DX 12 so my theory may be dead.

    Also, any chance its a windows 11 issue?

    Riboflavin on
  • BahamutZEROBahamutZERO Registered User regular
    doing a driver purge and reinstall is a good idea, use this utility just in case there's some glitched out cruft in the installation that upgrading or normal uninstalling doesn't catch https://www.guru3d.com/download/display-driver-uninstaller-download/

    BahamutZERO.gif
  • RiboflavinRiboflavin Registered User regular
    Rabbit hole.

    I downloaded windbg and looked at the .dmp files of the games that crashed that had .dmp files. I don't know if I am using it correctly as a caveat.

    I saw a access violation c0000005 error in each one. I tried googling it but it was less than helpful as it said it could be memory, voltage fluctuations, processor errors, video ram , it really didn't narrow down anything.

    I am going to reseat the card and wipe/reload drivers on Saturday and test the system RAM Sunday as I have a frustration level that won't let me work on it too long.

    I finished Baldur's Gate 3 and it worked fine and was stable. My kid told me it crashes every ten minutes now. It's weird. Dying light 2 works fine since I switched to dx12 but every other game has random crashes.

  • wunderbarwunderbar What Have I Done? Registered User regular
    Don't reseat the card. Just do the driver cleanup. I am more than 75% sure that's it.

    XBL: thewunderbar PSN: thewunderbar NNID: thewunderbar Steam: wunderbar87 Twitter: wunderbar
  • RiboflavinRiboflavin Registered User regular
    Did the driver cleanup in safe mode to no avail. Still crashed.

    Watched a video on underclocking a GPU, downloaded MSI afterburner, underclocked the GPU. No change.

    Finally, contacted MSI, Got an RMA, and am sending the card back under warranty. I am terrified my pc is going to be down months and they are going to send me the same card back and tell me they couldn't find anything.

    Hopefully my pessimistic view is faulty and they'll send me a new or functioning refurb but I've read some horror stories. Never read the internet. :-)

  • wunderbarwunderbar What Have I Done? Registered User regular
    Riboflavin wrote: »
    Did the driver cleanup in safe mode to no avail. Still crashed.

    Watched a video on underclocking a GPU, downloaded MSI afterburner, underclocked the GPU. No change.

    Finally, contacted MSI, Got an RMA, and am sending the card back under warranty. I am terrified my pc is going to be down months and they are going to send me the same card back and tell me they couldn't find anything.

    Hopefully my pessimistic view is faulty and they'll send me a new or functioning refurb but I've read some horror stories. Never read the internet. :-)

    RMA's don't work like that. You'll get a new card. They don't get your card and try to fix it and send it back.

    XBL: thewunderbar PSN: thewunderbar NNID: thewunderbar Steam: wunderbar87 Twitter: wunderbar
  • RiboflavinRiboflavin Registered User regular
    wunderbar wrote: »
    Riboflavin wrote: »
    Did the driver cleanup in safe mode to no avail. Still crashed.

    Watched a video on underclocking a GPU, downloaded MSI afterburner, underclocked the GPU. No change.

    Finally, contacted MSI, Got an RMA, and am sending the card back under warranty. I am terrified my pc is going to be down months and they are going to send me the same card back and tell me they couldn't find anything.

    Hopefully my pessimistic view is faulty and they'll send me a new or functioning refurb but I've read some horror stories. Never read the internet. :-)

    RMA's don't work like that. You'll get a new card. They don't get your card and try to fix it and send it back.

    This is what triggered my anxiety:

    https://www.reddit.com/r/MSI_Gaming/comments/qdhe7x/my_msi_repair_rma_nightmare/

  • BahamutZEROBahamutZERO Registered User regular
    Hopefully they don't give you the runaround like that.

    BahamutZERO.gif
  • wunderbarwunderbar What Have I Done? Registered User regular
    Riboflavin wrote: »
    wunderbar wrote: »
    Riboflavin wrote: »
    Did the driver cleanup in safe mode to no avail. Still crashed.

    Watched a video on underclocking a GPU, downloaded MSI afterburner, underclocked the GPU. No change.

    Finally, contacted MSI, Got an RMA, and am sending the card back under warranty. I am terrified my pc is going to be down months and they are going to send me the same card back and tell me they couldn't find anything.

    Hopefully my pessimistic view is faulty and they'll send me a new or functioning refurb but I've read some horror stories. Never read the internet. :-)

    RMA's don't work like that. You'll get a new card. They don't get your card and try to fix it and send it back.

    This is what triggered my anxiety:

    https://www.reddit.com/r/MSI_Gaming/comments/qdhe7x/my_msi_repair_rma_nightmare/

    Remember, that is one story. People usually don't come to the internet to say "you know, I had an issue, and my experience getting it dealt with was extremely positive and smooth and there were no issues."

    Most posts will be complaints and issues, where the vast majority of people just go about their day and don't feel the need to scream to the world "I had a good experience"

    XBL: thewunderbar PSN: thewunderbar NNID: thewunderbar Steam: wunderbar87 Twitter: wunderbar
  • RiboflavinRiboflavin Registered User regular
    Hate to keep dragging this post but when I underclocked with Afterburner will my replacement card be underclocked? Are the settings in the registry or are they somehow stored on the card? Will I need to go in and change it back when I get my replacement?

  • Phoenix-DPhoenix-D Registered User regular
    Riboflavin wrote: »
    Hate to keep dragging this post but when I underclocked with Afterburner will my replacement card be underclocked? Are the settings in the registry or are they somehow stored on the card? Will I need to go in and change it back when I get my replacement?

    If you're getting a new card, do a clean install. That will wipe all the previous settings.

  • RiboflavinRiboflavin Registered User regular
    Run the DDU again right? How does the DDU delete Afterburner settings?

  • BahamutZEROBahamutZERO Registered User regular
    edited February 7
    I am not sure if Afterburner is dumb enough to keep applying saved overclock profiles to a new card or not, but if it is, just uninstall Afterburner, and to be safe delete any files in its installation folder if they remain (default path is probably C:\Program Files (x86)\MSI Afterburner\ ). Specifically the internet tells me it saves your settings in ...\MSI Afterburner\Profiles\[a file name with a bunch of numbers and letters]. Then you can reinstall it and it should be back to default values. The way Afterburner works, overclocking changes you make with it only persist until reboot, and then have to be reapplied by running Afterburner again, it's not saved on the GPU or in the registry or anything that lasts between boots. If Afterburner is set to run automatically on Windows startup and you have it applying an unstable overclock that prevents you from going into its settings to turn that off, you could theoretically have to start Windows in safe mode so it doesn't launch extra stuff on startup, and uninstall it there; but I doubt that will be an issue for you.

    BahamutZERO on
    BahamutZERO.gif
  • RiboflavinRiboflavin Registered User regular
    Had a slight issue with no tracking number on RMA when they sent it back. Contacted MSI. It will be here tomorrow.

    A little worried because it is the same card. I asked what was done to it and was told the notes only said a component was replaced but not what component. I'll get it back to its cozy house in the case tomorrow and stress test it with tons of gaming this weekend. Wish me luck.

  • RiboflavinRiboflavin Registered User regular
    Popped the card back in and did the delete and reinstall drivers two step. It didn't take long before the card started messing up again. I got a bit ticked.

    Contacted MSI support, was a bit testy and misunderstood what they were telling me when they were doing the right thing. Felt real bad about it. Gave them a glowing review trying to fix my guilt.

    Sending the card back again. They sent me a label so I don't have to pay to ship it again and they marked it "replace don't repair". Don't know if I am getting a new one or a refurb but hopefully whatever they send I won't be doing this again. Got a little scared when I asked and they said "We'll have to see what is in stock." but I will try to not let my anxiety crank.

  • RiboflavinRiboflavin Registered User regular
    Ok, rabbit hole.

    Got a refurbed card from MSI.
    Had the exact same issue. Runs fine on desktop crashes in all games.

    1. Ran microsoft's memory checker again. Memory good according to it. Don't know how accurate it is.
    2. Bought a new power supply, spent 30 min replacing power supply. Still crashing.
    3. Bit the bullet and used my "play money" to order a low end 4060.Should be here Friday. Hoping MSI sent me a refurb with the same issue (not likely unless they have a common problem they don't want to address)

    If the 4060 crashes as well it leaves, memory(memtests say its good), processor and motherboard.

    ***What route would you go if this is your pc?


    It boggles my mind that the desktop is stable but I can't run a benchmark in cyberpunk without it crashing. I have 386 hours in cyberpunk before this started. Same with Baldurs gate. 100's of hours but can't play it more than 10 min now. Crashing on spell effects or camera rotation. 7daystodie crashes on fire effects (sometimes) Dying light 2 random crashes. Kids say Elden Ring crashes randomly but I don't play it. I have zero crashes watching youtube or working in open office. Its just games.

  • wunderbarwunderbar What Have I Done? Registered User regular
    Power Supply issue maybe.

    XBL: thewunderbar PSN: thewunderbar NNID: thewunderbar Steam: wunderbar87 Twitter: wunderbar
  • RiboflavinRiboflavin Registered User regular
    edited March 13
    wunderbar wrote: »
    Power Supply issue maybe.

    2. Bought a new power supply, spent 30 min replacing power supply. Still crashing. (Replaced my 1000watt no name with a 750 corsair didn't seem to change anything)

    Between the video card and power suplly I'm about 400$ in repairing a 1200$ pc.

    Riboflavin on
  • tsmvengytsmvengy Registered User regular
    What are your whole specs?

    Do you have PSU plugs attached to both the GPU (it probably requires 2) and the CPU (4 or 8-pin on the motherboard)?

    Have you tried backing up what you need and then doing a repair install of Windows?

    steam_sig.png
  • RiboflavinRiboflavin Registered User regular
    edited March 14
    This machine worked for a year. Built Jan 2023. Started crashing in games in January 2024.

    MSI GeForce rtx 3060 12gb Ventus 2x 12gb OC
    Gigabyte z790 atx motherboard
    Corsair 32gb (2x16) ddr5 4800
    Intel i7-13700kf
    coolermaster i70c lga1700 low profile fan
    I have 3 ssd's of different brands. A 250 boot drive and a 500 and a TB game drives.


    I now have a corsair rm750e I replaced a generic 1000 watt PSU.

    I have not tried repairing windows.

    Riboflavin on
  • tsmvengytsmvengy Registered User regular
    I would try a repair install of Windows and if that doesn't work, a full wipe and reinstall. Is your boot drive SATA or NVME?

    Also, I don't know that this is your issue, but you are using a $25 cooler designed to cool 65W processors on a $350+ processor with a TDP of 125W that can pull up to 250. You are probably CPU throttling A LOT. Especially if that cooler has some dust buildup on it.

    You would do a lot better with something like this, if it will fit in your case: https://www.amazon.com/dp/B0CL8Z43L8. What case do you have? We can check if it will fit.

    Not trying to pile on but just speaking frankly, I think your build could probably do with a re-assessment based on some guides to setting up a PC - e.g. do you have XMP enabled on your RAM?

    steam_sig.png
  • RiboflavinRiboflavin Registered User regular
    edited March 14
    Yeah, I bought a massive cooler which I still have. It was too wide and hit part of the MB so I couldn't use it so I looked for a highly rated low profile one. I will likely start tracking temps before replacing the MB but again, it worked for a year and I have periodically checked temps as part of this shenanigans. In fact temp tracking is probably the next step. IS there any tracking programs that will write a log? Not real familiar.

    I had the deepcool ag400. It hit like a cover or heatsink on the MB and wouldn't fit. I don't know the exact dimensions.


    ***I have changed my mind. I think I'm just going to go ahead and replace the fan anyway.***** The fan that didn't fit was 4.92 X 3.62 X 5.91 I will try to do some approximate measurements tonight

    Riboflavin on
  • tsmvengytsmvengy Registered User regular
    Riboflavin wrote: »
    Yeah, I bought a massive cooler which I still have. It was too wide and hit part of the MB so I couldn't use it so I looked for a highly rated low profile one. I will likely start tracking temps before replacing the MB but again, it worked for a year and I have periodically checked temps as part of this shenanigans. In fact temp tracking is probably the next step. IS there any tracking programs that will write a log? Not real familiar.

    I had the deepcool ag400. It hit like a cover or heatsink on the MB and wouldn't fit. I don't know the exact dimensions.

    I would be surprised if the cooler did not fit if you have an ATX or mATX board. They are designed to have a clearance zone around the CPU and coolers are designed to fit in that zone. The cooler should be oriented so the fan blows horizontally, not vertically (when the motherboard is in the case). Maybe it needed to be rotated 90 degrees one way or another.

    Temp tracking - HWinfo: https://www.hwinfo.com/

    Install it, go to Sensors, and click logging start, and it will save a CSV file and log your temps while you run your games.

    If everything worked fine for a year and now does not, then my suggestion is backup important docs and full wipe and Windows reinstall. There may be some driver or some update that is borked and figuring out exactly what it is.

    steam_sig.png
  • SiliconStewSiliconStew Registered User regular
    Another thing that is an easy and free test. Find out which RAM slot is the primary and run the system with just one stick of RAM in that slot. See if it still crashes. If it does, switch RAM sticks, again only using the primary slot, and see if it still crashes. While you mention passing a memory test, I've had similar random crashes in games but it was fine just running Windows that was down to one bad stick of RAM in a pair. It also passed mem tests which led me astray on troubleshooting.

    Just remember that half the people you meet are below average intelligence.
  • RiboflavinRiboflavin Registered User regular
    tsmvengy wrote: »
    Riboflavin wrote: »
    Yeah, I bought a massive cooler which I still have. It was too wide and hit part of the MB so I couldn't use it so I looked for a highly rated low profile one. I will likely start tracking temps before replacing the MB but again, it worked for a year and I have periodically checked temps as part of this shenanigans. In fact temp tracking is probably the next step. IS there any tracking programs that will write a log? Not real familiar.

    I had the deepcool ag400. It hit like a cover or heatsink on the MB and wouldn't fit. I don't know the exact dimensions.

    I would be surprised if the cooler did not fit if you have an ATX or mATX board. They are designed to have a clearance zone around the CPU and coolers are designed to fit in that zone. The cooler should be oriented so the fan blows horizontally, not vertically (when the motherboard is in the case). Maybe it needed to be rotated 90 degrees one way or another.

    Temp tracking - HWinfo: https://www.hwinfo.com/

    Install it, go to Sensors, and click logging start, and it will save a CSV file and log your temps while you run your games.

    If everything worked fine for a year and now does not, then my suggestion is backup important docs and full wipe and Windows reinstall. There may be some driver or some update that is borked and figuring out exactly what it is.

    When you say horizontally do you mean that airflow should be from mem slots to usb ports on back or should blow towards pcie slots?

  • tsmvengytsmvengy Registered User regular
    Riboflavin wrote: »
    tsmvengy wrote: »
    Riboflavin wrote: »
    Yeah, I bought a massive cooler which I still have. It was too wide and hit part of the MB so I couldn't use it so I looked for a highly rated low profile one. I will likely start tracking temps before replacing the MB but again, it worked for a year and I have periodically checked temps as part of this shenanigans. In fact temp tracking is probably the next step. IS there any tracking programs that will write a log? Not real familiar.

    I had the deepcool ag400. It hit like a cover or heatsink on the MB and wouldn't fit. I don't know the exact dimensions.

    I would be surprised if the cooler did not fit if you have an ATX or mATX board. They are designed to have a clearance zone around the CPU and coolers are designed to fit in that zone. The cooler should be oriented so the fan blows horizontally, not vertically (when the motherboard is in the case). Maybe it needed to be rotated 90 degrees one way or another.

    Temp tracking - HWinfo: https://www.hwinfo.com/

    Install it, go to Sensors, and click logging start, and it will save a CSV file and log your temps while you run your games.

    If everything worked fine for a year and now does not, then my suggestion is backup important docs and full wipe and Windows reinstall. There may be some driver or some update that is borked and figuring out exactly what it is.

    When you say horizontally do you mean that airflow should be from mem slots to usb ports on back or should blow towards pcie slots?

    The first one - airflow should be from the memory slots to the USB and other ports on the back.

    steam_sig.png
  • RiboflavinRiboflavin Registered User regular
    TLDR:It was likely the FAN and overheating.

    I have been building computers since the early 90's. The most problems I ever had was with a bad Slot A processor and I remember files and buffers for the config.sys.
    This led to Hubris and then to my downfall of spending $400 on a powersupply and 4060 trying to fix this.
    I did check temps but not under load. So a game would crash, I would open Gigabyte Control Center then go to fan and the temps would be in the normal range. I think the fan had enough time to bring the temp down.
    The fan worked for a year and temps were good so it had to be video card.
    MSI sent me a refurb because I insisted. Sorry MSI.
    When I started , waaaaaaaaaaay too late, to check temps in game I noticed 7daystodie getting up to 80-85C. I assume it crashed when I had many particle effects on the screen because it pushed the temp past a threshold.
    I don't know what caused the other fan to lose efficiency. Maybe dust on the sink or fan or a breakdown of thermal grease or whatever. Temps never got bad on desktop.

    I replaced the fan with a Darkrock PX4. Temps in game are about 20C lower across the board. I was able to boot cyberpunk, run a benchmark and load a game which I haven't been able to do before. Played Dyinglight 2 for about an hour with no crash. Ran Baldur's gate 3 about 15 min with no crash. I still need to test more but I am about 80% sure this was the issue. My games have not crashed since I replaced it but I haven't had time to play a ton but its definitely better.

    Thanks tsmvengy for fan shaming me and everyone else who made suggestions I may or may not have listened to. Now I need to figure out what to do with a refurbed 3060. :-)

  • BahamutZEROBahamutZERO Registered User regular
    edited March 19
    What CPU is it? 85 should be normal operating temperature for modern CPUs, and they are just supposed to slow themselves down if they hit 90c rather than crash, but if installing a better CPU cooler has alleviated the problems I guess that seems like the culprit. Fingers crossed that was all it was.

    BahamutZERO on
    BahamutZERO.gif
  • RiboflavinRiboflavin Registered User regular
    edited March 19
    What CPU is it? 85 should be normal operating temperature for modern CPUs, and they are just supposed to slow themselves down if they hit 90c rather than crash, but if installing a better CPU cooler has alleviated the problems I guess that seems like the culprit. Fingers crossed that was all it was.

    I don't think 85 was the final destination. The game I checked it on(7daystodie) was still running crashed at random but it was hard to get a temp exactly when it crashed. Also an old, not graphically intense game. I think the temp climbed slowly until it crashed. More demanding games climbed faster until it crashed. I definitely could not run a cyberpunk benchmark before swapping fans. It crashed 100% of the time. Runs now.

    I do find it interesting that it didn't get above 61C in 7days after the new fan.

    More importantly I can run Forbidden West when it comes out. ;-)



    Riboflavin on
Sign In or Register to comment.