-
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.
Both RX Vega 64 & 56 Beat GTX 1080 Ti in Forza 7 DX12 benchmarks
- Thread starter RollaC
- Start date
Caporegime
There's one main thread (Some other games do this too) but there's varying load spread amongst the other cores.
Also, driving games are simple. Take Dirt 2 when it first came out as the first DX11 title and how it ran etc.
Also, driving games are simple. Take Dirt 2 when it first came out as the first DX11 title and how it ran etc.
And this is the likey reason the Nvidia performance is so poor at the moment because it is such a bodged DX12 implementation that the load is not being properly shared across cores. Nvidia GPUs are getting starved which is why increasing resolution makes minimal impact on Nvidia GPUs.
Caporegime
We could do with seeing GPU usage.
Soldato
What Martini1991 says, just because it's using one core max doesn't mean the other cores are left doing nothing.
Caporegime
Soldato
AMD gpus need even more cpu grunt to function than the Nvidia ones though and they are doing fine here.
Actually it's probably down to the superior dx12 handling on AMD cards. They will be getting more out of what cpu reserves are on tap.
Last edited:
Soldato
Why would amd be any different then? If the CPU isn't feeling the GPU enough on Nvidia it won't be on AMD either.
Ryzen, benchmark results tell a story and that story is the game is using all the cores available. We know this because Ryzen is beating i7 7700k
If the game was only using the one core like you saying Ryzen wouldn't beat out Intel because Intel have better single thread performance.
Excuse after excuse...
The difference in this game is simply down to AMD having better DirectX 12 hardware.
That the only sensible conclusion to come isn't it. Can a big Nvidia driver gain back so much performance or will Nvidia have to throw lots of DX11 hardware at the DX12 problem?
Permabanned
Settings are all a bit bizarre in the demo. Won't let me select anything over 60 fps target....
Not really true. NVidia DX drivers intercept the API calls and and create multi-theaded command lists. They also do a lot of prepossessing of the shaders to maximize performance. If OFrza7 is using a DX12 engine on a single core then the NVidia GPUs might lag behind massively. AMD DX11 drivers have never been so sophisticated which is why they saw gains in Dx12. The reason AMD does well here is mostly a reflection of how bad the Nvidia drivers are handling the game engine, which is obvious when looking at the results.
AMD's drivers are very different to Nvidia drivers. If Nvidia's drivers are designed around heavy multi-threading and AMd's are not then NVidia will see massive issues and AMD woudln't.
Even if this conjectureis wrong, it doesn't really change anything. teh results are compeltely flawed, it doesn't take a rocket scientists to see that the numbers just don't add up.
You have absolutely zero understanding of DX12
You have a to make an argument from a fixed point. But let's ignore you're abrasiveness and please enlighten everyone. You seem to have intimate knowledge of AMD and Nvidia drivers. Please explain the differences. Can you also link the driver code base please.
These are the basic facts (number form compterbase.de):
at 1080p the 1080ti is only about 6% faster than the standard 1080. There is normally at least a 20% differnce and the ti is CPU-bound at this res in many games. At a 6% delta it is clear that the Ti is completely CPU bound.
At 1080P, thte Ti has almost the same min-frame rate as the RX570. Again, classic case of the Ti being CPU bound
When switching to 4K, suddenly the Ti take a commanding lead when the GPU is less CPU bound.
The ti compared to 1080 is now 20.5% faster, which is closer to what we see in other games but still a little lower (should be nearer 35%).
Comparing 4K to 1080p, the Ti slowed down only 18% despite rendering 8x as many pixels. The RX64 on the other hand slowed down by 40% Quite clearly, the Ti slowed down much less as the graphical load increased.
Everything points to Nvidia's drivers not working correctly with this title and/or soemhting wrong with the game engine.
at 1080p the 1080ti is only about 6% faster than the standard 1080. There is normally at least a 20% differnce and the ti is CPU-bound at this res in many games. At a 6% delta it is clear that the Ti is completely CPU bound.
At 1080P, thte Ti has almost the same min-frame rate as the RX570. Again, classic case of the Ti being CPU bound
When switching to 4K, suddenly the Ti take a commanding lead when the GPU is less CPU bound.
The ti compared to 1080 is now 20.5% faster, which is closer to what we see in other games but still a little lower (should be nearer 35%).
Comparing 4K to 1080p, the Ti slowed down only 18% despite rendering 8x as many pixels. The RX64 on the other hand slowed down by 40% Quite clearly, the Ti slowed down much less as the graphical load increased.
Everything points to Nvidia's drivers not working correctly with this title and/or soemhting wrong with the game engine.
No no let's talks drivers. What is the Nvidia driver doing wrong?
The adults are
Soldato
Lol everytime AMD wins its always the same. Something must be wrong with the drivers or the games broken.