I ran Memtest overnight from a USB stick & revealed no errors (I've never used it before, I just left it running & came back to a 'press escape to exit' message).
I have a retail copy of Win10 home, it normally automatically updates but I have been checking to make sure it's up to date over the past few weeks, and have checked (onboard Realtek) sound, network, monitor drivers etc are up to date, nvidia drivers are 398.82, the latest release according to geforce experience (again they normally automatically update).
I have the latest bios for my mobo, issued to correct for the recent intel hardware exploits, this was installed prior to getting the new GPU.
I have reseated the memory a couple of times since this problem started, first time I realised I had originally mounted the two 8gb DDR4 2133 sticks in slots 2 & 4 so tried moving them to 1 & 3, then when the issue persisted re-seated them to make sure.
The mobo was overclocked to 102%, and setting this back to 100% improves the situation, the new GPU automatically clocks itself with the EVGA Precision XOC app, subjectively disabling this on startup also seemed to improve the situation. I'm not getting CTDs on every station instance now, but it is still happening. Prior to installing this new card I had never seen this problem before, although I am aware others have had this or a similar issue for years.
Since replacing the PSU I've had no hard resets, so it may be that my old PSU just wasn't providing stable enough power when both the CPU & new power hungry GPU are being taxed. The only monitoring tools I have for the PSU simply show the voltages & they were easily within 5% tolerance ATX spec, they probably varied by less than 1-2% that I saw, the new one barely changes so I've stopped monitoring that.
I'll try underclocking the card today & muck about with various monitor combos & see if I can get it to crash with only the 144Hz monitor, and only a 60Hz monitor but framerate uncapped etc.
I appreciate your suggestions chaps, I am starting to run out of ideas.
I have a retail copy of Win10 home, it normally automatically updates but I have been checking to make sure it's up to date over the past few weeks, and have checked (onboard Realtek) sound, network, monitor drivers etc are up to date, nvidia drivers are 398.82, the latest release according to geforce experience (again they normally automatically update).
I have the latest bios for my mobo, issued to correct for the recent intel hardware exploits, this was installed prior to getting the new GPU.
I have reseated the memory a couple of times since this problem started, first time I realised I had originally mounted the two 8gb DDR4 2133 sticks in slots 2 & 4 so tried moving them to 1 & 3, then when the issue persisted re-seated them to make sure.
The mobo was overclocked to 102%, and setting this back to 100% improves the situation, the new GPU automatically clocks itself with the EVGA Precision XOC app, subjectively disabling this on startup also seemed to improve the situation. I'm not getting CTDs on every station instance now, but it is still happening. Prior to installing this new card I had never seen this problem before, although I am aware others have had this or a similar issue for years.
Since replacing the PSU I've had no hard resets, so it may be that my old PSU just wasn't providing stable enough power when both the CPU & new power hungry GPU are being taxed. The only monitoring tools I have for the PSU simply show the voltages & they were easily within 5% tolerance ATX spec, they probably varied by less than 1-2% that I saw, the new one barely changes so I've stopped monitoring that.
I'll try underclocking the card today & muck about with various monitor combos & see if I can get it to crash with only the 144Hz monitor, and only a 60Hz monitor but framerate uncapped etc.
I appreciate your suggestions chaps, I am starting to run out of ideas.
Last edited:
