This is where it’s nice to play on console. It’s going to work each and every time, from year 1 through year 10, the same way, each and every time.
Granted I have a GTX 1070, 16 RAM, i7 laptop but I don’t have overwatch for it. Just my PS4 Pro.
1 Like
Something is definitely up. I play on a 9700K/2070 and typically sit north of 240 FPS on medium/high settings.
Do you have local reflections on?
I doubt the problem is my in game settings everything is low or off, but thank you for your help
You’re right. If you’re actually using the “low” preset it’ll turn all that off.
Do you know what your GPU/CPU utilization looks like while playing?
About 15% GPU 40% CPU Utilization
The thing is, expect of extreme cases of of differences between the CPU and the GPU, the X component which is bottlenecking is not the 100% limiter of the performance.
For example, upgrading either the GPU(to 2080TI for example) or the CPU(to Ryzen 3700 for example), will give him some performance boost, just not close to the full potential they could have. It is probably true tho that the boost from the GPU probably will be the lowest in case of low settings, out of those 2 options.
The 2700x is super strong CPU, not crazy for gaming but i certainly not think it’s 100% bottleneck even on low settings.
I might assuming wrong tho, i didn’t checked it personally and i’m not sure if there is any documentation of extreme case like this when you are lowering every graphic settings to low and than comparing the CPUs.
How do you measure those GPU usage?
I found a large differences between what the Windows shows, and from what other GPU software showed me(MSI afterburner), not sure of the reason.
Or maybe it’s just issue with my Windows GPU monitoring, idk.
I was using task manager but I will check with msi afterburner
i think a fresh install of the nvidia drivers then go into your nvidia settings and make sure they are focused on performance.
1 Like
The GPU typically runs in parallel with CPU work (they are not serialized), so this is typically not true. I’m oversimplifying this quite a bit, but in general the cost to render a frame in a video game is commonly max(CpuTime, GpuTime), and not the sum of the two. As a result you typically only see an improvement when you upgrade the component that is bottlenecked. Someone running a 2070 on low settings will almost certainly see almost zero gains from using a lower render scale.
The 2700x is also a fairly modest CPU compared to other high end CPUs. It’s by no means a bad CPU but isn’t really even in the same class as a 9800K or a 9900K (or even the new 3700X, for that matter).
Update: I dropped my render scale to 75% and started using afterburner. My average Fps is a little higher now around 190-200 and afterburner put my GPU usage at around 40% according to afterburner
Something sounds out of whack, I get mid hundreds @ 1440P on a mobile RTX 2060 with tweaked medium settings. The desktop RTX 2070 is way more powerful.
If you’re playing in windowed mode, try changing to full screen.
Rtx 2070 at 1080p will play 300+ fps at all times on ultra.
1080p will be 300+ fps.
1440p will be 144-300 fps on ultra (My gtx 1080 runs that.)
4k 70fps - 150 fps. Ultra.
With the way the game runs your gonna want 2k 300fps. Higher resolution for clearer distance aiming, the higher fps reduces input lag.
1440p at high or very high should net that at all times.
I’m pretty sure there is some sum value that you can gain in most cases.
Even if it’s just 5% better.
Can’t say that fore sure for this exact setup, but i have exactly one class below that in GPU(RTX2060) and CPU(2600X) and i do gain a lot from 75% render scale(low settings), with a big fps boost.
I didn’t test that extensively tho(in terms of the dips in big fights) because usually i capping it on 144~150fps to match my monitor.
Last time I tried this on my 1080Ti I got literally no improvement from 75% render scale, as I expected. I immediately changed it back because I dont like the quality decrease (I have no issues with seeing enemy outlines).
And what is your CPU?
I’ll check also with friend of mine who also have 1080TI and see what fps he gets compare to 100% render scale.
I’m pretty sure he set that to 75% and it actually helped him a lot with the FPS and the SIM value specifically. And he doesn’t have crazy CPU, i7 6700, not k.
It also possible that it was just because bad OW patch or specifically issue in his PC.
If the drivers are up to date and not corrupt or causing any issues (for example pascal and anything later than 417.xx will result in worse fps, especially 425.xx and 430.xx drivers), and you also set up your Windows correctly by disabling game bar and game mode and all that other bloatware junk, then I would like to ask, did you enable XMP in the bios as well as to tweak some other settings? Ryzen highly benefits from high clocked ram, you might have 3000mhz ram, but does it actually run at 3000mhz? Also, while Zen as an architecture is not weak, Zen Gen1 certainly lacks in SINGLE core performance, OW is one of those games that favours single core performance and high clocked ram over everything else, more cores and a good graphics card doesn’t do much, especially the latter if the settings are set on low with 75% render scale which automatically shift most of the work to the cpu in order to maintain high fps and creates a bottleneck there, as the gpu has basically nothing to do.
I have an i7 4770K which is for sure the bottleneck on my machine.
I’m not saying it never will, I’m saying it sometimes will. But playing Overwatch on all low settings on a very high end GPU like a 1080Ti or a 2070/2080… you’re barely utilizing it.
Does your friend play on all low settings, or medium/high/etc?
Sure it does’t give you the full performance, but would you say that if you had only GTX1080 that would give you the same performance because of this CPU bottleneck? I think you would gain something, even on low settings.
Not sure, it was quit some time ago and he played with the settings a lot.
Yes I would. I dont think it would change a thing.