Amazing! It seems that when it was set to 3840x1600p with dlss in quality without the frame generation and without rt on, it seems that the image quality even with dlss was as close as a native resolution 3440x1440p and the performance jump ranged from 120 to 140 fps, more pixels, more sharpen and better image quality with 3840x1600p + dlss quality than 3440x1440p + dlss quality.
Yep I agree. Definitely sharp. And something to consider if you wanted to go to a 38" 3840x1600. While using dlss you'll likely have a pretty sharp picture. And the performance is pretty close to the 3440x1440!
@@kgtech007 So cold. Your 4080 is so cool during all the tests, i was blown away by the thermal capacity your model has, really impressive, it must be an amazing build! Is your in the watercooler block?
@@joaovitornepomuceno5290 thanks! Ya it runs great! The GPU is air cooled and the CPU is on a 240 rad. I'm going to do a quick video on it soon. I think partly it has to do with the case I have and also the custom fan curve for the case and the GPU/CPU. It's all helps.
Nice vid, but can you enable frametime graph and also the power draw of the video card? I'm running a 3070 FE on 3440x1400, to get absolutely free performance you can try to undervolt -50mv and try to increase the GPU frequency until it is not stable + set your power limit to maximum, in some cases it will draw the same as no undervolt+overclock, but in my case I am getting about 7-10% of performance for free
Thanks! Sure I can do that. Ive been having trouble with MSI afterburner not showing some stats on the overlay. But I have done some over clocking and testing with power draw differences between settings like having dlls on and off with raytracing. I haven't under volted my 4080 yet. It's so efficient and cool I'm not sure I'll mess with it. Even over clocked till it's unstable it runs cool.
![5 САМЫХ ОШИБОЧНЫХ РЕШЕНИЙ В CYBERPUNK 2077 [by Evoice Erebus x Kupikod]](http://i.ytimg.com/vi/ZXLALPduZqo/mqdefault.jpg)
Excited for this new upload!
Nice video, there is no ultrawide benchmarks of pathtracing anywhere else
Thank you! I'm working on another with some overclocking aswell. I'm also working on another ultrawide path tracing video.
Amazing! It seems that when it was set to 3840x1600p with dlss in quality without the frame generation and without rt on, it seems that the image quality even with dlss was as close as a native resolution 3440x1440p and the performance jump ranged from 120 to 140 fps, more pixels, more sharpen and better image quality with 3840x1600p + dlss quality than 3440x1440p + dlss quality.
Yep I agree. Definitely sharp. And something to consider if you wanted to go to a 38" 3840x1600. While using dlss you'll likely have a pretty sharp picture. And the performance is pretty close to the 3440x1440!
@@kgtech007
So cold. Your 4080 is so cool during all the tests, i was blown away by the thermal capacity your model has, really impressive, it must be an amazing build!
Is your in the watercooler block?
@@joaovitornepomuceno5290 thanks! Ya it runs great! The GPU is air cooled and the CPU is on a 240 rad. I'm going to do a quick video on it soon. I think partly it has to do with the case I have and also the custom fan curve for the case and the GPU/CPU. It's all helps.
Nice vid, but can you enable frametime graph and also the power draw of the video card?
I'm running a 3070 FE on 3440x1400, to get absolutely free performance you can try to undervolt -50mv and try to increase the GPU frequency until it is not stable + set your power limit to maximum, in some cases it will draw the same as no undervolt+overclock, but in my case I am getting about 7-10% of performance for free
Thanks! Sure I can do that. Ive been having trouble with MSI afterburner not showing some stats on the overlay. But I have done some over clocking and testing with power draw differences between settings like having dlls on and off with raytracing. I haven't under volted my 4080 yet. It's so efficient and cool I'm not sure I'll mess with it. Even over clocked till it's unstable it runs cool.