Unigine Valley - ±130FPS not more

[Khimari]

Hi guys,

I just wanted to ask, why is it that when I run Unigine Valley benchmark, I run everything at lowest settings

yet when benchmark runs I get only around 130FPS?

Yes, it jumps between 40 to 160, but the average is still around 120-130. When I run old 3DMark03, I can get around 4000FPS in 1024x768.

I mean, I'm ok with Unigine, no complains, I'm just curious why the minimum and maximum setup status doesn't produce too much different FPS. I can run it maxed at allmost same FPS.

To me it is strange that quality and resolution does not matter. BTW when running the minimum setup, CPU and GPU are both used only at 30-34%. Why is it so? It's the same with mostly newest games, at low setup everything runs not as fast as old stuff. Setting GPU power management to maximum performance does not help. Overall system setup is always set to maximum performance.

Does someone have an idea? Bad drivers? Software framerate locked? I know that for example Counter Strike: Source is locked to max of 300FPS. Can this be the same?

CPU: i5-4570 @ 3.20GHz
GPU: ROG STRIX-GTX1080-O8G-GAMING
RAM: 16 GB DIMM A-DATA DDR3-1600

 

[weberdarren97]

Valley Benchmark does rely heavily on Tessellation, which breaks down a polygon into many smaller polygons in order to produce smoother edges. This process can sometimes produce millions of extra polygons for the GPU to render, causing fps to drop. That's likely the culprit.

 

[weberdarren97]

4000 fps seems unreasonable? I don't think it's accurate. Are there even any types of display cables capable of carrying information that quickly? I don't think there are.

It may be that you simply have your 1080 overclocked too high... Above 2000MHz is risky territory. It may be malfunctioning at that speed.

 

[Khimari]

4000 fps seems unreasonable? I don't think it's accurate. Are there even any types of display cables capable of carrying information that quickly? I don't think there are.

It may be that you simply have your 1080 overclocked too high... Above 2000MHz is risky territory. It may be malfunctioning at that speed.

Hm, maybe. Well I'm not touching my GPU, it does all it's bussiness on it's own so I do not need to care, but I'm not asking about 3DMark03, it's old software and all old software runs at such speeds. 3DMark03 can produce around, dunno, 180K points or 160K on complete benchmark, I don't remember clearly, it's not important. I'm more curious about the Unigine benchmark. Can the low FPS difference be caused by geometry itself?

 

[weberdarren97]

Valley Benchmark does rely heavily on Tessellation, which breaks down a polygon into many smaller polygons in order to produce smoother edges. This process can sometimes produce millions of extra polygons for the GPU to render, causing fps to drop. That's likely the culprit.

 

[Khimari]

Valley Benchmark does rely heavily on Tessellation, which breaks down a polygon into many smaller polygons in order to produce smoother edges. This process can sometimes produce millions of extra polygons for the GPU to render, causing fps to drop. That's likely the culprit.

Fair enough, I thought the same.

EDIT:

But why the system is used only at 30%?

 

[weberdarren97]

But why the system is used only at 30%?

Frame rate targeting.

I know, why put framerate targeting into a benchmark utility? That's why I no longer use Valley.

 

[Khimari]

But why the system is used only at 30%?

Frame rate targeting.

I know, why put framerate targeting into a benchmark utility? That's why I no longer use Valley.

Well it looks like this. At least that's what the overall feel and taste is. If this is true it's quite boring...