> people actually believed Oxides meme about DX12 async compute giving 30% performance improvement on AMD hardware
TOPKEK
O
P
K
E
K
>>51791273
>DX12 worked how it was designed to be implemented
>nVidia kicked up shit because their hardware wasn't optimised for the workload
>Threw money at it
>DX12 ded
Why laugh, when you should cry. We're stuck with shit GPU performance and no advancements for another decade.
>>51791273
> nvidia dx11 faster than amd dx12
> in an amd sponsored game
my sides
>>51791349
>nVidia DX12 slower than nVidia DX11
Oh, we're not even getting started on the laughs yet.
>DX12 is the future, seriously it will bring a whole new level of performance guys!! It's like 50% better on existing hardware!!!11!
>3-4 fps on a DX12 specifically developed game.
I should just get a new hobbie and ditch video games for ever.
>>51791362
> nvidia dx12 slower than nvidia dx11, but still faster than amd dx12
loving every laugh
>>51791273
>DX12 async compute giving 30% performance improvement
You mean the feature that the developers were forced by Nvidia to scale back significantly just so it wouldn't make Nvidia's DX12 support look as shit as it is?
There were news everywhere that Oxide was pressured by Nvidia to reduce implementing async compute. A while after that, behold, Nvidia's performance is not shit anymore.
It's both hilarious and sad that you think it's a good thing that everyone gets less features so that your favorite, awful GPU company don't look as bad as it should.
>>51791273
I believe dx12 was Microsoft's way of sparking up some competition in the gpu market, if amd didn't fail so fucking hard and surpassed nvidia thanks to dx12 (like they said they would), nvidia would be forced to either lower their prices a lot in the future, or to really step up their innovation game.
>>51791413
>slower nvidia dx12 still faster than amd dx11&dx12
>>51791428
Source on Oxide scaling back their use of async?
All I heard is that they added a special-case hack for nVidia systems which disables async completely as it reduced performance, but they were still using it on AMD systems.
>>51791386
>A game that according to it's developers doesn't even utilize a feature of a new API that heavily
>When the only OS which supports that API hasn't even even been out for 6 months
>Somehow a surprise that the advantage of the new API isn't that big when compared to an API that's been out for 6 years already
Then again I do tend to forget that /g/ pretty much is a sub-board of /v/ and incapable of doing anything beyond looking at numbers... Like many other genuinely technical things, the concept of driver maturity seems to be beyond what our /v/ spillover seems to be capable of understanding.
