Absolutely. The demo is not about RTX. It is about agnostic Ray tracing. Meaning no special Api are required. And no it’s not half of the RTX card being off.
Now will the new Cryengine likely have RTX api in use for the next engine? I am sure it will.
Did you know that if the RTX is not doing Ray Tracing the extra chips are just doing nothing? Where as the Amd Asic(?) Can & do other functions.
This demo is about agnostic ray tracing or Open Ray Tracing functionality.
Microsoft DXR and Chronos Vulkan raytracing API are both GPU agnostic but make use of RT cores on GPU’s that have them. Crytek have ignored industru standard API’s in order to deliberately NOT make use of any GPU that might have ray tracing hardware.
Their demo still runs faster in cards that do have RT cores, even though they’ve deliberately disabled them when they could have ALSO optionally used them IF they are present.
Deliberately disabling a co-processor when using it would be a trivial if then else statement is not “progress”, or clever.
As far as I know, Crytek uses the standard DX12 raytracing calls. What hardware is being used is up to the driver. Unlikely that Nvidia doesn’t use their RT cores in DX12 raytracing calls.
The graphics in the Crytek demo is actually a lot better than the RTX implementation in many games.
Back to the actual topic: Other headsets seem to work very well with AMD GPUs. Pimax doesn’t. I don’t know whose fault this is, but it’s unlikely that AMD decided to purposely not make Pimax work, so it must be Pimax doing something different than other headsets.
At very least Pimax could regularly inform us what they are doing to fix the issues. Even if it was just something like: “This issue is caused by AMD’s driver. We reached out to AMD 1 month ago and asked for the current status. AMD replied they are still working on it.”
I doubt Pimax is actually doing much, because we’re not getting any updates from Pimax for over a year now and I have not gotten a confirmation from AMD that they are aware of these issues.
Nope, this demo is on DX11, read the article.
Nvidia is not about open standards. Let’s take Nvidia issues to it’s own topic. Vulkan & gpu open are open standards. Dx & Nvidia api are licensed closed standards. People often confuse MP3 as an open standard as well.
As @Willyfisch said lets get back on topic.
Real3D for life! (later renamed to Realsoft3D - presumably to avoid trouble with some richer people who appared and took the name for themselves :P)
I hopped over to download the Neon Noir demo, and give it a go, but that apparently took signing up for an account at Crytek, even though it’s free, so duck that. 
Everyone these days seems to want sign ups. Lol
There’s nothing really new or revolutionary about RTX cores, they just are built to process data in a way that Nvidia considered it preferable and simplified to compute RT faster than the regular method (at least strictly for gaming purposes…), where in reality it is simply a matter of using plain floating point ray tracing calculations, and AMD compute units are and always have been better at doing it in the regular way exactly like most professional ray trace software is doing, and Crytek just wanted to show this…in other words, Nvidia just tried to circumvent the problem by engineering a simplified RT rendering queue, but such simplified ray tracing method is just as easily done in real-time using the AMD FP hardware pipeline 
If you look at the high end scientific, medical and military applications, most of them are using AMD hardware rather than Nvidia solutions, because it is way more efficient in doing so.
By the same logic there isn’t really anything new or revolutionary about GPUs. GPUs are just built to do highly parallel floating point processing more efficiently than CPUs. I was running simple Ray tracing demos on my CPU at 30 FPS in 720p resolutions four year ago.
It is great that real time Ray tracing can be done on AMD hardware that isn’t specialized and can run non-ray tracing work loads as well.
“Real-time” is not a well defined comparison. One could be rendering 1080p at 60fps while the other could be doing 1080p at 100 fps. Both are real time.
A demo that compared the performance of the non specialized hardware to specialized hardware for the same raytracing workload would be more interesting. We can’t really say that it is just as easily done with non specialized hardware as with specialized unless the performance for a specific workload is performed. (As mentioned it is also easy to rayteace on CPU but slower)
Specialized hardware may not be necessary for real time Ray tracing but that doesn’t necessarily mean that it is not more efficient. It is difficult to tell without a fair comparison.
Even if it turns out that the non specialized hardware is slower than specialized hardware. It is great though that AMD can run raytracing workloads in real time as it will provide a cheaper alternative to Nvidia.