Ashes of the Singularity Founders Benchmark Now Available

I seem to be one of the folks who get worse performance with DX12 over DX11, by about 5 to 15%.

== Hardware Configuration =================================
GPU:        NVIDIA GeForce GTX 770
CPU:        GenuineIntel
               Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz
Physical Cores:            4
Logical Cores:            4
Physical Memory:         8503021568 bytes

How long does the benchmark run? It ran for like 40 minutes then I couldn't find a way to end it other than alt f4.

How did you start the Benchmark?  From the main menu?  It should only take 3 minutes.

I guess the point of these benchmarks is to show the difference between few things on the screen (low batches) and large number of things on the screen (high batches)

In my 1080p medium benchmarks, both the DX11 and DX12 versions run at 72fps at "low batch". However, at "high batch"  the DX11 version drops by 33fps whereas the DX12 version drops by 21fps.. and I think this is what Microsoft is keen to highlight about DX12

Summary:

DX12 at UHD (4K) - Crazy setting  - Difference between low + high:  21% (5fps) Avg: 20.4 fps 

DX12 at UHD (4K) - Crazy setting  - AVG: 20.4 FPS

And then DX12 at 1080p - Crazy setting - AVG: 33.8 FPS

DX11 at UHD (4K) - Crazy setting - AVG: 21.1 FPS

DX11 at 1080p - Crazy setting - AVG: 32.3 FPS

DX12 at UHD (4K) - Medium setting - AVG: 36.7 FPS

DX12 at 1080p - Medium setting - AVG: 60.2 FPS

DX11 at UHD (4K) - Medium setting - AVG: 37.5 FPS

DX11 at 1080p - Medium setting - AVG: 53.2 FPS
 

Yep, that's currently pretty typical with Nvidia cards. I lost about 10% with a GTX 970 and the benchmark articles released earlier this week (or was it last week?) indicate the same.

Ok never mind dummy was using the map benchmark. lol

Do the results upload anywhere?

It says compare on the leaderboards. Is that real?

SLI work? (i'm going to guess no - i have 4 titan Xs)

i'm getting about 29fps @ 4k with crazy

i cant seem to run the game on dx12.... and yes im on windows 10. i try to launch it from my library in steam..it starts then it closes...i tried going to my steam folder on my hard drive..and launching dx12 but it says dx11 for the benchmark...solutions anyone? or am i doing doing something wrong?

Nope. and using alternate frame rendering in the nVidia settings causes a world of screen flicker.

EDIT: Try the AFRGPU option which enables alternate frame rendering

As with others, I also have worse performance under DirectX 12.

Intel i7 920 OC'd to 4.0ghz, 12GB DDR3, EVGA Geforce 970 FTW 2.0 w/ additional 125mhz OC (total of 1341 core boost block)

On average, I'm about 2-5fps LESS on DirectX 12 vs 11.

DX11:

DX12:

Hopefully there will be optimizations over the next few months. I can understand not seeing a huge improvement but seeing a loss in performance is pretty egregious.

Try disabling MFAA in the Nvidia Control Panel.

With it disabled, dxdiag will show only DirectX 11.3 support, but once disabled, it should allow for DirectX 12 to be utilized. You may or may not have to reboot after disabling it, can't remember.

 I think this bug 'may' of been fixed w/ the latest Nvidia driver tho.

Thanks man! it was already off but game profile wasn't added to it.

ok it loads directly from my library but when i click to run benchmark my driver crashes .

My first post like Founder lifetime

I do not know how upload pictures or where is the leaderboard so 

DX11 at 1080p - Crazy setting on 980 Ti

Average Framerate

1. 33.9 fps (29.5 ms)
2. 44 fps  (22.7ms)
3. 35 fps (28.6ms)
4. 27 fps (37.1ms)

here is my heavy results with same settings FX-8350/HD 7970

DX11

Total Time: 57.735985
Avg Framerate : 91.936279 ms (10.877098 FPS)
Weighted Framerate : 95.895103 ms (10.428061 FPS)
Average Batches per frame: 29422.806641

DX12

Total Time: 57.871265
Avg Framerate : 51.670776 ms (19.353300 FPS)
Weighted Framerate : 52.491089 ms (19.050854 FPS)
CPU frame rate (estimated framerate if not GPU bound): 29.191771 ms (34.256229 FPS)
Percent GPU Bound: 99.766823%
Driver throughput (Batches per ms): 3174.050293
Average Batches per frame: 33720.179688

Configuration
API: DirectX
Resolution: 1920x1080
Fullscreen: True
Bloom Quality: High
PointLight Quality: High
Glare Quality: High
Shading Samples: 16
Terrain Shading Samples: 8
Shadow Quality: High
Temporal AA Duration: 6
Temporal AA Time Slice: 2
Multisample Anti-Aliasing: 1
Texture Rank : 1

i think it works... 

At the moment there is a problem with Nvidia cards and anti-aliasing using DirectX 12 (MSAA and Temporal AA). With activated anti-aliasing DX12 is slower as DX11. Not sure if the Nvidia driver or the engine is responsible for this bug...

not a bug. Nvidia's dx11 drivers optimize this perf. Their dx12 ones do not yet.

http://www.pcper.com/reviews/Graphics-Cards/DX12-GPU-and-CPU-Performance-Tested-Ashes-Singularity-Benchmark/Results-Avera

Looks like AMD CPUs are really bad in this benchmarks.

i3 4330 beating FX 8370. Is this just a bug or true performance of FX CPUs?

I would love to know this myself.

When the initial review benchmarks were on the net. I thought there must have been something wrong. So as my gaming rig is the same spec, FX-8370 + R9 290X. Figured I would check it out, sure enough, the results were almost identical. Checking the CPU and core utilization, it showed that the CPU was maxing out quite a bit. So the only question is, why is the game hammering an AMD FX so much and causing such a significant bottleneck to the GPU?

All I can think of is that there is a problem with the compiler, or that it isn't utilizing all the advanced instruction sets that the Vishera cores support.

Hmm. 

Got FX 6300 and R9 270X .... my performance would be very low. So if I want to run(buy) this game I will have to upgrade my GPU (R9 290). Then CPU to i5 or i7?... worth or not?

Holy moly, do not go out and make some expensive impulse purchases. Let the dust settle a bit. On my i5 2500k and GTX 770, hardware worse than yours (I think....I don't keep up on my AMD hardware generations), I was getting just over 30 fps averages in the benchmark, pegging the CPU. That is hardly very low performance. In my alpha gameplay games so far FPS rarely dipped down that far, so the game is quite playable on my modest hardware.

One of the major advantages of dx12 is to properly utilize all cores, isn't it? Then how can a i3 4330 beat an FX 8370? I tested it with my oc'd fx8350 and r9 290 and got very similar results myself as well.

his 270x should be significantly slower than a 770. A 290 would be beneficial to him

How? If Fx 6300 is slower than i3.

We see benhcmarks and I see that FX 8350 is losing against i3. 

Be cautios when looking at numbers. They may be reporting average frame rates, which isn't very useful. It's possible there are some scenes where the I3 is at 80fps for a while which makes it look like it's faster.

The number to look at is the heavy score - which is where the CPU is really under major load. It's like a pickup truck, what you really care about is how fast you can go when you've got your bed filled with stuff, not how fast you can go when it's empty.