- Jan 28, 2018
- 17
- 0
- 1,510
I'm back again with another 970 question... I've been running a modded bios on my card for overclocking, I have raised the maximum voltage to 1.31mv and raised the power limit to 287watts.
It allows me to apply a overclock on the core around 110 to 150+ offset which gets me around 1580-1590mhz.
(I have the MSI GTX 970 Gaming 4G) (I put this thread in the wrong spot, sorry)
Most of the time, it's completely stable. It doesn't have any artifacts when playing games, temps do hit the 70s on intense scenes only when the card starts to use 250+watts. But the thing here is, these clocks are stable on a few games, but on fallout4, for example, it crashes randomly. I could be playing for a half hour, maybe even 2 hours, and it'll randomly crash to desktop without any errors/logs. Meanwhile, the game runs completely fine. No stuttering as well... Erm... until l run out of ram and hit the pagefile, but the game runs fine without any graphical bugs.
The only way I got around this, was to lower the clocks to 1530mhz but it still crashed. It crashed quicker than before but the crashes were not consistent. They were more random this time around.
Once I set the core clock to 1520mhz, the crashes stopped on that game. But its really annoying as I lose 15-10fps if I lower those clocks, as I'm playing my games maxed out at 3840x1620. I get around 45-60fps depending on the location and game.
In bf1(3,4), metro, ATS, project zomboid, hitman, rising storm, witcher etc. is stable at 1585mhz or 1575mhz along with a +600oc on the memory 4130mhz. I can even play metro last light at 1595mhz, but 1590 is more stable in that game.
I mean, does my card need more voltage? Power? Is the temp causing the crashes? Or is it the Nvidia drivers, crapping themselves? Maybe its the Vrms getting too warm? I can't really tell how warm the Vrms is because my GPU doesn't have sensors for them. Unless there's a way to see the temps that I don't know. I already use HWINFO to see the power consumption, temps, voltage etc. And MSI afterburner for overclocking.
The back side of the card isn't hot, I mean its warm pretty warm at load but it isn't burning hot. The highest temp I've seen my card hit was 76c. And that doesn't happen often. My card usually hangs around 65c to 70c on full load. Of course, when it's using max power, it'll hit 70 to 75c but it rarely even uses 260watts.
I just don't understand why my games crash just so randomly. I mean its working fine, nothing wrong then poof your game crashed to desktop.
It doesn't skip frames, there are no artifacts, the temps are normal it isn't throttling and my CPU is essentially bottlenecking my GPU too.
I only have an i7 2600 and 8gbs of ram in single channel mode, as the other 2 banks don't work. My GPU usage sometimes spikes up and down in some games with my CPU being around 50 to 70% utilization. Its mostly the reason why I play on *4k* at ultra settings. It lowered the CPU bottleneck and made the game look better. Some games give me a dx11 error, which I know is from overclocking but its really stupid how my games run fine. Completely fine. Then all of a sudden the driver or game doesn't like it and crashes to desktop with or without an error. It's seriously frustrating and infuriating.
I owned my 970 since I bought it brand new. I only modded it to improve performance and I do not regret doing that. My games wouldn't have been anywhere near playable at 4k on stock settings. I practically gained 25-30fps on some occasions with these settings. Especially at 1440p, some games would get 100fps with some settings on high/medium. When I modded my card, I was able to turn up all of the settings to the max while getting much more stable framerate and no more power/voltage throttling. It was a near 30-40fps gain over the stock, nonmodded un-overclocked settings.
If anyone has any reason or a way to get higher clocks without these random crashes, that would be greatly appreciated. If anyone wants to look at my current bios file for my card, I'll gladly upload it.
It allows me to apply a overclock on the core around 110 to 150+ offset which gets me around 1580-1590mhz.
(I have the MSI GTX 970 Gaming 4G) (I put this thread in the wrong spot, sorry)
Most of the time, it's completely stable. It doesn't have any artifacts when playing games, temps do hit the 70s on intense scenes only when the card starts to use 250+watts. But the thing here is, these clocks are stable on a few games, but on fallout4, for example, it crashes randomly. I could be playing for a half hour, maybe even 2 hours, and it'll randomly crash to desktop without any errors/logs. Meanwhile, the game runs completely fine. No stuttering as well... Erm... until l run out of ram and hit the pagefile, but the game runs fine without any graphical bugs.
The only way I got around this, was to lower the clocks to 1530mhz but it still crashed. It crashed quicker than before but the crashes were not consistent. They were more random this time around.
Once I set the core clock to 1520mhz, the crashes stopped on that game. But its really annoying as I lose 15-10fps if I lower those clocks, as I'm playing my games maxed out at 3840x1620. I get around 45-60fps depending on the location and game.
In bf1(3,4), metro, ATS, project zomboid, hitman, rising storm, witcher etc. is stable at 1585mhz or 1575mhz along with a +600oc on the memory 4130mhz. I can even play metro last light at 1595mhz, but 1590 is more stable in that game.
I mean, does my card need more voltage? Power? Is the temp causing the crashes? Or is it the Nvidia drivers, crapping themselves? Maybe its the Vrms getting too warm? I can't really tell how warm the Vrms is because my GPU doesn't have sensors for them. Unless there's a way to see the temps that I don't know. I already use HWINFO to see the power consumption, temps, voltage etc. And MSI afterburner for overclocking.
The back side of the card isn't hot, I mean its warm pretty warm at load but it isn't burning hot. The highest temp I've seen my card hit was 76c. And that doesn't happen often. My card usually hangs around 65c to 70c on full load. Of course, when it's using max power, it'll hit 70 to 75c but it rarely even uses 260watts.
I just don't understand why my games crash just so randomly. I mean its working fine, nothing wrong then poof your game crashed to desktop.
It doesn't skip frames, there are no artifacts, the temps are normal it isn't throttling and my CPU is essentially bottlenecking my GPU too.
I only have an i7 2600 and 8gbs of ram in single channel mode, as the other 2 banks don't work. My GPU usage sometimes spikes up and down in some games with my CPU being around 50 to 70% utilization. Its mostly the reason why I play on *4k* at ultra settings. It lowered the CPU bottleneck and made the game look better. Some games give me a dx11 error, which I know is from overclocking but its really stupid how my games run fine. Completely fine. Then all of a sudden the driver or game doesn't like it and crashes to desktop with or without an error. It's seriously frustrating and infuriating.
I owned my 970 since I bought it brand new. I only modded it to improve performance and I do not regret doing that. My games wouldn't have been anywhere near playable at 4k on stock settings. I practically gained 25-30fps on some occasions with these settings. Especially at 1440p, some games would get 100fps with some settings on high/medium. When I modded my card, I was able to turn up all of the settings to the max while getting much more stable framerate and no more power/voltage throttling. It was a near 30-40fps gain over the stock, nonmodded un-overclocked settings.
If anyone has any reason or a way to get higher clocks without these random crashes, that would be greatly appreciated. If anyone wants to look at my current bios file for my card, I'll gladly upload it.
Facebook
Twitter
Instagram