• 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.

NVIDIA Titan V Raytraces Battelfield V in RTX Mode at Proper Perf (but does not have any RT cores).

Caporegime
Joined
8 Jul 2003
Posts
30,063
Location
In a house
It is a bit of a remarkable story really, but users have enabled RTX mode on a Nvidia Titan V, which works quite well and performs as fast as the RTX 2080 Ti. Titan V, however, is Volta, and Volta does not have any RT cores.

Of course, Volta is a very powerful card with loads of headroom and tensor cores. DX-R mode enabled, in theory, can work on any compatible DirectX 12 graphics card as long as it is fast enough, as the DirectX Raytracing is merely an extension to the DX12 API. If the hardware is not supported, it could run it in a software modus. However, the findings raise the question, are dedicated RT in hardware cores actually needed?

In the example, the Titan V with a Volta GV100 GPU according to users in this thread over at 3dcenter, would produce framerates that are almost equal to those of the GeForce RTX 2080 Ti. One user reports 69 fps in Ultra HD resolution at set Ultra high with both cards. Both graphics cards have been overclocked and cooled with water cooling with equal framerates and image quality. Yet another user reports he was playing in a 60-100 fps at 1440p ultra with 80 fps average. In our own review, the RTX 2080 Ti with DXR raytracing averaged 84.3 fps which is pretty damn close towards our results as tested in our Battlefield V article.

These new findings do make us wonder what the actual effect of RT has on performance. It has to be stated, these are merely user reports and we cannot verify any test methodologies used here (testing in an RT enhanced environment would be one concern that comes to mind).

Regardless, these are interesting find but for now, please do take them with caution and a few grains of salt. Below two settings screenshot the users used, as well as a screengrab with RTX enabled on the Volta card.

https://www.guru3d.com/news_story/n...er_perf_(but_does_not_have_any_rt_cores).html
 
Titan V is £2800 but yes it is a pointless exercise.

For the record RTX was first demonstrated on Volta not Turing.


You sure? I know dice were using volta cards up until late on to implement it, but the first demos at the launch event were on turing.
 
Anyone got a good comparison video of bf v with and without Ray tracing?

Been playing it a fair bit and for some reason it started crashing if using DX12 version too often, I tried with or without DXR same issue so have gone back to DX11 for now. What I can tell you is it looks noticeably worse in my opinion, much preferred the quality with DXR enabled so once the holidays are done i'll mes around to see if its working more reliably on again...
 
First demo was done before Turing launch on Volta.


So it would seem that "rt cores" are a gimmick and are more than likely reprogrammed cuda cores? If that's the case even pascal is capable of rtx effects if its more of a software based feature.
 
So it would seem that "rt cores" are a gimmick

No - RT cores can do specific types of instruction very efficiently they also have a different memory configuration to enable this. I'm not sure what is going on in this case but the implementation in BF V is a hybrid reflection system that uses ray casting and a lot of tricks (that are run on conventional cores) and isn't fully leveraging the capabilities of RT cores and was developed initially using Volta hardware so its possible the Volta specific paths are still there for it.
 
Back
Top Bottom