2070 super underperforming?

[Snakedudo]

So I’ve recently got a new pc with a 2070 super, and according to benchmarks, on siege I should be getting high 200 FPS. Apparently even 1660 ti get around 160 FPS. Both are benchmarked on 1080 and ultra.I’m hovering around 130 only on high settings, and 1080p, so could it be a system problem or gpu problem?

 

specs

ryzen 5 3600x

msi 2070s ventus oc

msi b450i gaming plus ac

8gb*2 3600 g.skill ripjaws 

1tb intel 660p ssd

corsair sf600 plat 

[5x5]

Updated drivers, BIOS, firmware, etc? XMP enabled? No power saving features in Nvidia control panel?

[Vasllo]

I often don't get the results I see on benchmarks, weirdly enough. But is your memory actually set to "3600MHz"? And what's the CPU on these benchmarks? Odds are it's your CPU capping your GPU, which you can easily check with  MSI Afterburner's overlay for GPU usage while you play.

[5x5]

2 minutes ago, Vasllo said:

I often don't get the results I see on benchmarks, weirdly enough. But is your memory actually set to "3600MHz"? And what's the CPU on these benchmarks? Odds are it's your CPU capping your GPU, which you can easily check with  MSI Afterburner's overlay for GPU usage while you play.

He needs CPU usage for that, if both the CPU and GPU aren't at 100% then it's a software issue. Honestly, unless he's messed up, a 3600 won't limit a 2070S

[Vasllo]

8 minutes ago, 5x5 said:

He needs CPU usage for that, if both the CPU and GPU aren't at 100% then it's a software issue. Honestly, unless he's messed up, a 3600 won't limit a 2070S

CPU usage on Ryzen often doesn't mean anything, because it usually bottlenecks on latency or something else. On the other hand, if you check the GPU usage, you can be sure whether it is being bottlenecked (<90% usage) or not. And saying an R5 3600 won't bottleneck an RTX 2070S as blank statement is rather incorrect, because it all depends on what he's trying to do. If he's playing on ultra at 4k, sure, if he's playing an e-sports title (like he is) at FHD resolution, then it's not so obvious anymore. I can see the R5 3600 being the bottleneck here, Intel still has a wide lead in some titles - and was probably the CPU used on the benchmark he saw -, and not long ago, the R5 2600 struggled to keep 120fps on average.

[5x5]

4 minutes ago, Vasllo said:

CPU usage on Ryzen often doesn't mean anything, because it usually bottlenecks on latency or something else. On the other hand, if you check the GPU usage, you can be sure whether it is being bottlenecked (<90% usage) or not. And saying an R5 3600 won't bottleneck an RTX 2070S as blank statement is rather incorrect, because it all depends on what he's trying to do. If he's playing on ultra at 4k, sure, if he's playing an e-sports title (like he is) at FHD resolution, then it's not so obvious anymore. I can see the R5 3600 being the bottleneck here, Intel still has a wide lead in some titles - and was probably the CPU used on the benchmark he saw -, and not long ago, the R5 2600 struggled to keep 120fps on average.

I'm not sure you're looking at the same CPUs. The Zen 2 CPUs are among the best in single and multi. Intel CPUs can barely lead by 4-5% in most cases. If the 3600 is a bottleneck, almost certainly every CPU is a bottleneck. If you're looking at LoL or CS then even the 9900K at 5GHz is a bottleneck. Latency isn't the issue here, it's the poorly optimized game engines or buggy VGA drivers. My card also sees suboptimal usage on an 8th gen i7, does that mean it's bottlenecked? Nope, it means the game engine is poorly optimized in most cases

[Vasllo]

2 minutes ago, 5x5 said:

I'm not sure you're looking at the same CPUs. The Zen 2 CPUs are among the best in single and multi. Intel CPUs can barely lead by 4-5% in most cases. If the 3600 is a bottleneck, almost certainly every CPU is a bottleneck. If you're looking at LoL or CS then even the 9900K at 5GHz is a bottleneck. Latency isn't the issue here, it's the poorly optimized game engines or buggy VGA drivers. My card also sees suboptimal usage on an 8th gen i7, does that mean it's bottlenecked? Nope, it means the game engine is poorly optimized in most cases

If he's seeing it performing better on benchmarks, yes, it does mean something on his PC is bottlenecking. If the benchmarker can run at higher FPS, it's not the software optimization. Also not likely to be driver, but he can try to DDU it just for peace of mind. Intel has a bigger than 4-5% advantage in several games still, specially compared to the mid end R5. Compared to the i9 9900k, it will usually have lower framerates, so this might be what's causing the difference perceived by the OP.

[Vasllo]

15 minutes ago, 5x5 said:

I'm not sure you're looking at the same CPUs. The Zen 2 CPUs are among the best in single and multi. Intel CPUs can barely lead by 4-5% in most cases. If the 3600 is a bottleneck, almost certainly every CPU is a bottleneck. If you're looking at LoL or CS then even the 9900K at 5GHz is a bottleneck. Latency isn't the issue here, it's the poorly optimized game engines or buggy VGA drivers. My card also sees suboptimal usage on an 8th gen i7, does that mean it's bottlenecked? Nope, it means the game engine is poorly optimized in most cases

Here you can check the performance of the R5 3600 compared to Intel.

 

 

[Snakedudo]

8 hours ago, Vasllo said:

I often don't get the results I see on benchmarks, weirdly enough. But is your memory actually set to "3600MHz"? And what's the CPU on these benchmarks? Odds are it's your CPU capping your GPU, which you can easily check with  MSI Afterburner's overlay for GPU usage while you play.

Using ryzen master I turned my ram up to 3600 and my computer wouldn’t boot after that, so I cleared cmos

[Vasllo]

16 hours ago, Snakedudo said:

Using ryzen master I turned my ram up to 3600 and my computer wouldn’t boot after that, so I cleared cmos

Oh well, I've seen that before, try setting it in your BIOS, and if still doesn't work after training, you may increase the DRAM voltage to 1.35v, max is 1.5V for DDR4 AFAIK.