• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Ashes of the Singularity Coming, with DX12 Benchmark in thread.

'Nvidia mistakenly stated that there is a bug in the Ashes code regarding MSAA. By Sunday, we had verified that the issue is in their DirectX 12 driver. Unfortunately, this was not before they had told the media that Ashes has a buggy MSAA mode. More on that issue here. On top of that, the effect on their numbers is fairly inconsequential. As the HW vendor's DirectX 12 drivers mature, you will see DirectX 12 performance pull out ahead even further.'



We've offered to do the optimization for their DirectX 12 driver on the app side that is in line with what they had in their DirectX 11 driver. Though, it would be helpful if Nvidia quit shooting the messenger.

http://forums.oxidegames.com/470406
 
Last edited:
Nvidia are really trying to make themselves look stupid now. No doubt their fanboys customers will still defend them. Even more embarrassing that the dev even offered to sort Nvidia's own affairs out, just hilarious.
 
Last edited:
So now AMD finally have driver performance which is about the same as Nvidia, some people ( and NVidia) are getting worried. Maybe now we will see the true potential of AMD cards and all those extra SP's without the DX11 driver handicap. If Ashes is not a true representation of DX12 games then what is? This game performs tons of draw calls which is exactly what DX12 is designed to enable.

If newer games run good on AMD and Nvidia then it's better for eveyone since we have more choice and potentially better gpu prices.
 
Last edited:
Just for the record , i din't even mention the other gpu vendor's name. I am not here to discuss how hey perform but mention no DX 11 profile has been made so far to this bench/game. Now i go to sleep :D
 
In supreme commander the benchmark was never the exact same twice in a row and resolution didn't seem to change the result. Might explain why higher settings appeared to give better frames at times. In fact I don't think anything changed the result by much at all and if you ran it at 1024x768 it was practically the same as 1600x1200. Strange but its true.
 
In supreme commander the benchmark was never the exact same twice in a row and resolution didn't seem to change the result. Might explain why higher settings appeared to give better frames at times. In fact I don't think anything changed the result by much at all and if you ran it at 1024x768 it was practically the same as 1600x1200. Strange but its true.

The game was heavy CPU bound, i remember playing some great games with friends. When i had an Athlonx2 4200+ and the game time dilated we could take our merry time setting up defenses while an experimental trundled across the map towards us.

Then i upgraded to a 955BE and we were **** out so often because the game was much faster. XD

Still an awesome game though. On a scale no other game had before it in terms of units.
 
That is not how DX12 works at all, The developers have to write multi gpu support into their Rendering engine. All that happened with DX12 is that it allowed developers to do this themselves in the engine which is more efficient than the current bolted on top of DX11 method

Pretty much what I said :D Why go back and do it all over again.. surely that's DX11 thinking.
 
I wonder if results from other sources will show a similar pattern. Hopefully FM and Unigine will have something for us soon. Oxide being a bit of a AMD PR mouthpiece in recent times unfortunately throws an unavoidable level of suspicion on these results.

It was strange that the DX12 Star Swarm tests never got any further outings post that one Anandtech piece, I wonder why that was locked up tight.
 
I wonder if results from other sources will show a similar pattern. Hopefully FM and Unigine will have something for us soon. Oxide being a bit of a AMD PR mouthpiece in recent times unfortunately throws an unavoidable level of suspicion on these results.

It was strange that the DX12 Star Swarm tests never got any further outings post that one Anandtech piece, I wonder why that was locked up tight.

That thing was a ballake, it's a damn shame they didn't release an acceptable end user GUI, write in some vulcan and metal support and it could have been the de facto API benchmark. Now AOTS is probably going to garner an enhanced level of sales just so people can play with DX12 to see what it can do. So draw your own conclusions why Star Swarm was dropped but I have me a few ideas ;)
 
Dx12 looks like win to me. On high end CPUs it moves bottleneck to the GPU no matter how high end it is, which is exactly what you want as a gamer.

Not much else to be gleaned from this other than AMDs DX11 driver for this game sucks.
 
So much waffle so little knowledge.

On a single GPU I have not found a RTS that does not scale with resolution and I have played quite a few. There again that was on DX11, if this is not the case with DX12 there must be something wrong with the API.

What I would recommend you do is actually use some of the hardware you are talking about and also do some benching as you don't seem to know that good well written ones like Heaven, Firetrike, Vantage etc do scale with resolution. You only start seeing this failing with multi GPU setups or tiny resolutions.

Before people draw conclusions about DX12 we need to see results on some reliable benches and games. Also I think both the AMD and NVidia results look wrong so it is not a brand thing.


+1

More pixels = more work = lower FPS regardless of what API or bench you are using.:)



It's genuinely embarrassing to see you completely misread every single post I make then running around telling people to run crappy benchmarks that have zero relevance as proof of your superior knowledge.

Try reading what I said, IT'S AN RTS, not one of the benchmarks you names is an RTS. RTSs are CPU limited, not GPU limited, this has been the case in 9 out of 10 RTSs since forever. IF you are stuck at 30fps because the game thread(not the driver threads) are entirely limited on the CPU you frequently don't see scaling in RTS's as you go up in resolution because you aren't at your GPU limit at 1080p and might still not be at the GPU thread limit at 1440p or even potentially 4k.

I literally supplied a link showing exactly that, a RTS which showed a 0.1fps(more importantly, no proportional drop in performance) going from 1650 to 1920 resolution. In a FPS this would probably drop performance 20-30%, in an RTS performance didn't drop off at all.

For someone who insists they know what they are talking about, entirely ignores the link that proved you wrong, decided to say you know about RTSs so why don't I run non RTS benchmarks that have zero connection to RTS performance.

Well done, you told me to run benchmarks which do scale with resolution, I can name you 100 in game benchmarks from FPSs, RPGs and all other types of genre that scale with resolution as you'd expect. But THIS benchmark is an RTS and you are claiming it's troubled because it's not scaling with resolution and I've directly explained to you why this is relatively normal behaviour with RTSs.
 
All the performance increases will be in RTS and MMOs really, the improvements that we will get in existing FPS and other games will probably be small but we could push those games in ways we never could before. AOTS, alpha aside should pretty much be the absolute pinnacle of what DX12 and the other LL APIs can do.

Look at Mantle, average FPS gains in most titles, HUGE minimum FPS games in most of them. Thief is basically an FPS and gained a huge amount in minimum framerate from low level API. Games that are massively CPU bound will gain more, but there will be noticeable gains in most games if you look for it. Having played Thief and other games, it's smoother in Mantle, not all games were. It's a beta and most game support was effectively beta as well, it's there to be tested and give devs some experience for future games more than for a perfect final experience.

There are two effective ways DX12 will help you get better performance in games. The first is the more direct GPU one, better access to the hardware, better utilisation of the hardware, much more efficient draw calls with much lower overhead which simply means even if you are using the same number of draw calls it might save 2-3ms getting that data to the GPU. THat is the more direct gains. The indirect gains will be for MMO/RTSs, where if you can drop the driver overhead from say 30% to 5% you are indirectly giving the game thread 25% more available performance. AS these games are heavily CPU bound you effectively have a GPU that isn't being used as it waits for the game thread to finish, by speeding up the game threads the gpu ends up being used more efficiently.

The former should improve most games, the latter will help with any significantly CPU bound games which would be in general RTSs, MMOs and more complex simulators.
 
It's genuinely embarrassing to see you completely misread every single post I make then running around telling people to run crappy benchmarks that have zero relevance as proof of your superior knowledge.

Try reading what I said, IT'S AN RTS, not one of the benchmarks you names is an RTS. RTSs are CPU limited, not GPU limited, this has been the case in 9 out of 10 RTSs since forever. IF you are stuck at 30fps because the game thread(not the driver threads) are entirely limited on the CPU you frequently don't see scaling in RTS's as you go up in resolution because you aren't at your GPU limit at 1080p and might still not be at the GPU thread limit at 1440p or even potentially 4k.

I literally supplied a link showing exactly that, a RTS which showed a 0.1fps(more importantly, no proportional drop in performance) going from 1650 to 1920 resolution. In a FPS this would probably drop performance 20-30%, in an RTS performance didn't drop off at all.

For someone who insists they know what they are talking about, entirely ignores the link that proved you wrong, decided to say you know about RTSs so why don't I run non RTS benchmarks that have zero connection to RTS performance.

Well done, you told me to run benchmarks which do scale with resolution, I can name you 100 in game benchmarks from FPSs, RPGs and all other types of genre that scale with resolution as you'd expect. But THIS benchmark is an RTS and you are claiming it's troubled because it's not scaling with resolution and I've directly explained to you why this is relatively normal behaviour with RTSs.

Try using a decent CPU with your RTS games, it works for me lol.

And while we are on the subject you have more or less made my arguement for me.

This game is a crap benchmark for DX12 and should not be taken seriously.:D
 
I do love this place at times like this.

Game released, amd performance is gash, developers and nvidia are to blame.

This demo is released, nvidia performance is sub par (apparently), nvidia are to blame.

It's not like oxide have any previous for this kind of thing now is it.
 
I do love this place at times like this.

Game released, amd performance is gash, developers and nvidia are to blame.

This demo is released, nvidia performance is sub par (apparently), nvidia are to blame.

It's not like oxide have any previous for this kind of thing now is it.

All I want to see is a bench that demonstrates DX12 well, to me it does not really matter whether AMD or NVidia come out on top as it is DX12 that I want to see working.

As to all the arguing over drivers if people are unbiased they will realise that drivers from both brands will be substandard for some time to come as DX12 is very new.:)
 
'Nvidia mistakenly stated that there is a bug in the Ashes code regarding MSAA. By Sunday, we had verified that the issue is in their DirectX 12 driver. Unfortunately, this was not before they had told the media that Ashes has a buggy MSAA mode. More on that issue here. On top of that, the effect on their numbers is fairly inconsequential. As the HW vendor's DirectX 12 drivers mature, you will see DirectX 12 performance pull out ahead even further.'



We've offered to do the optimization for their DirectX 12 driver on the app side that is in line with what they had in their DirectX 11 driver. Though, it would be helpful if Nvidia quit shooting the messenger.

http://forums.oxidegames.com/470406

Good find there Humbug. :)
 
All I want to see is a bench that demonstrates DX12 well, to me it does not really matter whether AMD or NVidia come out on top as it is DX12 that I want to see working.

Looks working to me:

79b.jpg


79a.jpg
 
Back
Top Bottom