Post Your BaseMark GPU Scores!

[xkm1948]

because the GPU *ISN'T BEING UTILIZED* by this benchmark.

Exactly. Unoptimized OpenGL driver for this benchmark by RTG

 

[RealNeil]

You need to show cpu and gpuz

Like this?

So it seems that this test doesn't like SLI mode? A pair of 1070Ti cards should fare better than this, shouldn't they?

 

[Athlonite]

So even after an update that really didn't fix anything of great significance I still get an error when trying to run it in Vulcan and still see it lock up at 99% under OpenGL although the experience mode works fine in OpenGL but fails to even attempt to run in Vulcan mode so I've come to the conclusion that this benchmark app is poorly coded and a complete waste of time and probably shouldn't have been let out of the stable door until it worked properly on everything they seem to think it should work on

 

[BadFrog]

 

[defiiancecp]

Exactly. Unoptimized OpenGL driver for this benchmark by RTG

It's good that you're finally agreeing with me! I said in my very first post this appeared to be a problem with OpenGL implementation in the driver, to which you replied that it was not a problem.

 

[xkm1948]

It's good that you're finally agreeing with me! I said in my very first post this appeared to be a problem with OpenGL implementation in the driver, to which you replied that it was not a problem.

Well we have been agreeing on the same thing since the beginning regarding the problem: the driver.

It is just I dont see RTG trying to improve the driver for this specific benchmark. It takes money and human resources to improve OpenGL for GCN, per application. That is where the beauty of Vulkan and DX12 shine---no need to do per application optimization.

So yeah, OpenGL performance is kinda irrelevant as long as the application has proper Vulkan or DX12 implementation.

So even after an update that really didn't fix anything of great significance I still get an error when trying to run it in Vulcan and still see it lock up at 99% under OpenGL although the experience mode works fine in OpenGL but fails to even attempt to run in Vulcan mode so I've come to the conclusion that this benchmark app is poorly coded and a complete waste of time and probably shouldn't have been let out of the stable door until it worked properly on everything they seem to think it should work on

The problem maybe on both end. I did some digging with debugging tool. The problem seems to stem from both RTG's driver as well as the benchmark. The 18.6.1 driver has Vulkan 1.1.73 while this benchmark is still on Vulkan 1.0.

 

[AlwaysHope]

I ran this on my general purpose PC, but upon submission to BasemarkGPU website, it reports the version from here is out of date & thus not acceptable.

 

[iWalkingCorpse]

https://powerboard.basemark.com/submit/546645/result

2000-score is pixel 2
1100-score is BlackBerry Key2

 

[T4C Fantasy]

https://powerboard.basemark.com/submit/546645/result

2000-score is pixel 2
1100-score is BlackBerry Key2

this is awesome, we need more phones tested

 

[xkm1948]

Blackberry is still alive??? Holy shit.

Too bad they don’t have IOS or macOS version

 

[BadFrog]

Blackberry is still alive??? Holy shit.

Too bad they don’t have IOS or macOS version

iOS version is called Metal benchmark. Here’s my results on a iPhone 7 and then 7 plus

 

[T4C Fantasy]

iOS version is called Metal benchmark. Here’s my results on a iPhone 7 and then 7 plusView attachment 103016View attachment 103017

I need morr info say the SoC name and gpu name and which metal version is it
there is no min avg max score?
its hard to compare on the normal board unless it has the same setup

 

[Athlonite]

The problem maybe on both end. I did some digging with debugging tool. The problem seems to stem from both RTG's driver as well as the benchmark. The 18.6.1 driver has Vulkan 1.1.73 while this benchmark is still on Vulkan 1.0.

Perhaps we should send them the latest Vulcan SDK to use instead of them using old crap

 

[cadaveca]

Where's the AMD CPU results?

I guess everyone has the same problem I do with AMD CPU?

 

[Athlonite]

I'd put some up if only I could this dame bench to work properly

 

[xkm1948]

Where's the AMD CPU results?

I guess everyone has the same problem I do with AMD CPU?

You mean GPU right? There are multiple RyZen 2 CPU up there

 

[T4C Fantasy]

You mean GPU right? There are multiple RyZen 2 CPU up there

This thread has brought tpu so many new members xD

 

[xkm1948]

This thread has brought tpu so many new members xD

Really? I have no idea. People loves benchmark what can I say?

@theoneandonlymrk We need your EK water cooled Vega64 on there!

 

[RealNeil]

You mean GPU right? There are multiple RyZen 2 CPU up there

Some AMD CPUs have their own GPUs on board. Maybe he means them?

 

[TheBigNoob]

I'm not sure if I'm doing this right, but here it goes anyways (someone here suggested I post this)


https://powerboard.basemark.com/hwsubmit/21922/result

 

[T4C Fantasy]

I'm not sure if I'm doing this right, but here it goes anyways (someone here suggested I post this)


https://powerboard.basemark.com/hwsubmit/21922/result

nothing makes sense anymore

 

[cadaveca]

You mean GPU right? There are multiple RyZen 2 CPU up there

ryzen2, but no ryzen, and no TR. For me with TR, test finishes, but as it is generating score or whatever the app just disappears Shit software. Oh, I see some red tehre now, dunno how I missed it before...

 

[xkm1948]

I'm not sure if I'm doing this right, but here it goes anyways (someone here suggested I post this)


https://powerboard.basemark.com/hwsubmit/21922/result

nothing makes sense anymore

Comparing with the rest of the 1080Ti this makes no sense. Unless he has SLI on?

Another possibility is the internal timer bug. Some benchmarks uses system time stamp for calculating avg FPS, bugs in system time stamp will result in strange results. This once score is ~twice as much as a similar system.

I recommend pull that score again until TheBigNoob have another one that at least have minium FPS on there. As of now it just makes no sense at all

 

[mouacyk]

Look what I found:

Despite the claim it took two years to make, it's remarkable how fast obvious bugs have been found.

If you run the benchmark minimized it'll double your score. That's a terrible design.

Tl;dr would not recommend.

 

[T4C Fantasy]

Look what I found:

wow