Vulkan in Rainbow Six Siege Tested - A major boost for modern graphics cards

[WYP]

Vulkan is now the API of choice for users of new graphics cards.



Read more about Rainbow Six: Siege's Vulkan update and its performance consequences.

 

[NeverBackDown]

Just wanted to say. THANK YOU for including the GTX 1080 in the benchmark. You'd be surprised how many reviews go live these days without including this card despite it's popularity...


Other than that based off my 2 runs of DX11/Vulkan with my GTX 1080, I came to the same conclusion, but holy moly what a difference Vulkan makes for everything else!
We were CPU limited before and that 2080ti basically gained 30 FPS at 1440p. That's insane.

What is even more interesting, is Ubisoft said in their testing of Vulkan vs DX12, Vulkan offered better CPU performance. I'm not surprised with a 2080ti making that big of a jump. This is probably one of if not the best modern API implemnation in any game. Seems to be stable too.

I will stick to DX11 with my GTX 1080, just for sake of stability, but if they can continue to improve then I will switch over eventually.

 

[AngryGoldfish]

That's a huge performance jump. It makes me wonder why more developers don't use Vulkan. It seems to be consistently a good performer. There are a couple of outliers if I remember rightly, but generally it's been consistent for developers to eek out the most from their engine.

 

[Warchild]

Just wanted to say. THANK YOU for including the GTX 1080 in the benchmark. You'd be surprised how many reviews go live these days without including this card despite it's popularity...


Other than that based off my 2 runs of DX11/Vulkan with my GTX 1080, I came to the same conclusion, but holy moly what a difference Vulkan makes for everything else!
We were CPU limited before and that 2080ti basically gained 30 FPS at 1440p. That's insane.

What is even more interesting, is Ubisoft said in their testing of Vulkan vs DX12, Vulkan offered better CPU performance. I'm not surprised with a 2080ti making that big of a jump. This is probably one of if not the best modern API implemnation in any game. Seems to be stable too.

I will stick to DX11 with my GTX 1080, just for sake of stability, but if they can continue to improve then I will switch over eventually.

Thats seriously nuts. I'd expect a few % increase, but not double figures.

 

[NeverBackDown]

That's a huge performance jump. It makes me wonder why more developers don't use Vulkan. It seems to be consistently a good performer. There are a couple of outliers if I remember rightly, but generally it's been consistent for developers to eek out the most from their engine.

It takes a lot of work to rewrite an entire pipeline and getting the same stability, performance, etc. Lack of tools and support out there as well. However as time goes on it grows and therefore will gain popularity.

Thats seriously nuts. I'd expect a few % increase, but not double figures.

That's the joy of reducing CPU cycles for needless things. Speeds everything up considerably!

 

[Warchild]

im curious now how from a pure gaming perspective AMD holds up against Intel. Since Intel still has that edge.

Talking about if Vulcan was widely adopted.

 

[NeverBackDown]

im curious now how from a pure gaming perspective AMD holds up against Intel. Since Intel still has that edge.

Talking about if Vulcan was widely adopted.

Well if someone here has both a Zen 2 chip and a 9900k could run this test with a 2080ti at 1080p and see under DX11/Vulkan the differences and similarities(if any) I doubt have either, unfortunately.

Though DX12 and Vulkan are both mostly aimed at reducing CPU usage by reducing the required cycles required, that would aid AMD in IPC just because it has less to do, it also means Intel can go even faster too. In the end, the % delta may remain the same.

 

[ET3D]

Thanks for the article. It's the kind of thing I enjoy seeing.


NeverBackDown did raise an interesting point though. The 1080 showed hardly any change on your system, but apparently had a huge boost on a more CPU-limited system. Perhaps you could investigate this further?

 

[Peace Ð]

Just wanted to say. THANK YOU for including the GTX 1080 in the benchmark. You'd be surprised how many reviews go live these days without including this card despite it's popularity...

Would love to see the 1070 in there as well, but as it sits right between, there's not much to cover anyways. Seems like the 10 series is not affected as much anyways :/

 

[WYP]

Would love to see the 1070 in there as well, but as it sits right between, there's not much to cover anyways. Seems like the 10 series is not affected as much anyways :/

It comes down to time really. I have a GTX 1070 there, but I didn't want to test too many cards that feature the same architecture. We can approximate where cards like the GTX 1070 will lie.