I've just seen the actual test on YouTube and I'm not impressed tbh. Nothing is moving except the camera.
However, performance looks good for Nvidia RTX3090. I'm excited for the future when full RT games will be viable.
I've just seen the actual test on YouTube and I'm not impressed tbh. Nothing is moving except the camera.
However, performance looks good for Nvidia RTX3090. I'm excited for the future when full RT games will be viable.
That would be around 20 times raytracing performance than what we have today. Or it would require some other methods like voxel based lighting/cone tracing.
I got 18.76 or something like that with a 2070S. Pretty meagre results but I haven't been able to get an Ampere GPU and my next card will be (hopefully) 6800XT. Benchmarking fun will happen soon...
That would be around 20 times raytracing performance than what we have today. Or it would require some other methods like voxel based lighting/cone tracing.
Well, it means we'll live to see it happen! I can't wait.
I've just seen the actual test on YouTube and I'm not impressed tbh. Nothing is moving except the camera.
Agreed, and scene complexity is fairly flat. Since the scene is static there isn't any need to rebuild BVH for dynamic geometry which is what you would find in real-time games. We will need to keep in mind this benchmark definitely is not a measure of real-time game performance.
yep RT is anything but ready, imo Nvida use RT on 2xxx series as an excuse to start raising prices
Yes and no. Having dedicated cores and always pushing for bigger chips are one reason, the other was the lack of competition. Hopefully AMD will change that in 2021.
As for RT it's still too early: 3090 not getting even 60fps out of 1440p just shows we are at least 2 to 3 years away of real RT.
I've just seen the actual test on YouTube and I'm not impressed tbh. Nothing is moving except the camera.
However, performance looks good for Nvidia RTX3090. I'm excited for the future when full RT games will be viable.
You do realize that for a benchmark of rendering, motion of any kind is entirely irrelevant, right? As long as every testing pass is consistent across all machines it makes absolutely no difference what so ever whether a scene is static or dynamic. The rendering is what matters and as long as the test always renders the same things the same way, exactly what is being rendered does not matter.
Actually, a static scene is a far, far better test of rendering because it minimizes CPU involvement to practically zero and that is what you want from a test of rendering. Think about it: what does the animation? Hint - it's not the GPU.
With these results, this is obviously a very challenging test for today's hardware.
Yes and no. Having dedicated cores and always pushing for bigger chips are one reason, the other was the lack of competition. Hopefully AMD will change that in 2021.
As for RT it's still too early: 3090 not getting even 60fps out of 1440p just shows we are at least 2 to 3 years away of real RT.
like said RT is not ready if that test proves any thing it how heavy that RT is in a STATIC image full RT in gameisnt happen anytime soon the 3090 is only card that actual good performance and by that i mean the only card has "near" 60 fps and that card cost how much now?
like said RT is not ready if that test proves any thing it how heavy that RT is in a STATIC image full RT in game anytime soon the 3090 is only card that actual good performance and by that olny card has "near" 60 fps and that card cost how much now?
Not gona see anything in game anytime soon in true form
But you don't have to render an entire scene in RT in order to use RT, as evident by a number of games that use RT and get over 60fps on not a 3090.
So what your actually trying to say is, full scene RT is not ready - but that kind of invalidates your earlier point about the 2000 series because they were never really intended for full scene RT.
:D