At GTC, NVIDIA’s Don Palomo conference, they showcased the technologies that will enable apply Path Tracing in games, although so far they’ve only shown it for Cyberpunk. Obviously, since this is an event created by them and the marketing of the company is very powerful, we must separate the wheat from the chaff. And that’s what we’re going to do.
For some time now, Jen Hsen Huang’s company has chosen to implement new technologies not in terms of hardware, but to take advantage of the enormous computing power left by its graphics chips to do so using software. But not the one that runs on your CPU, but CUDA-based libraries and programs that run as compute shaders, i.e. for general purpose and not for manipulating graphics primitives, for solving problems which could be solved with fixed function accelerators. would do the job automatically, without developer involvement.
CUDA to link them all to NVIDIA
We have an example of this in Shader Execution Reordering, a function that NVIDIA introduced in its RTX 40 which is based on the possibility of rearranging the order of execution of the different threads that go to the GPU cores according to the execution units that are available at any time. This is a job that should be left to the kernel planner, but NVIDIA has decided to make it a feature of its own libraries.
That is, there is no hardware implementation that does this job and so on with a host of functions that are promoted recently. What they did was create a GPU, the AD102, and its graphics card, the RTX 4090, with so much power that no current game really uses it 100%, and then they launched a series of justifications in the form of proprietary software to justify its existence and bind developers to its ecosystem.
NVIDIA Path Tracing is not a standard
All this brings us to the question of the NVIDIA Path Tracing SDK and its big difference compared to Ray Tracing in 2018. At that time the initiative came from Microsoft to create a version of DirectX that allowed us to use what we let’s call it hybrid rendering, i.e. using ray-tracing algorithms to solve a series of visual problems such as reflections, transparency, light transport, etc., which are not possible otherwise. Therefore, a series of standards agreed by several manufacturers for the creation of a standard was born.
Now, although Path Tracing is a real-time 3D scene generation algorithm that’s been around for decades, what NVIDIA presented at the GTC isn’t, and it’s NVIDIA’s umpteenth attempt to tie users to your platform. . Like the brand’s DLSS, it is software that only runs on its graphics cards. While this brings benefits, it has made NVIDIA a de facto monopoly in both high performance computing and artificial intelligence, and it has been for years.
Just look at the technologies that make up the NVIDIA Path Tracing SDK, all proprietary and impossible to run on a competing GPU. Which are:
- DLS 3.
- RTX Direct Lighting (RTXDI).
- NVIDIA real-time denoisers (NRD).
- Opacity micro-map (OMM).
- Shader Execution Reordering (SER).
- Irradiance cache.
All are proprietary technologies not implemented in hardware but running on CUDA and requiring an RTX 4090 to give good performance.
A technology reserved for the privileged
Since game developers want to sell as many units as possible, it makes no sense to adopt a technology that only 0.31% of users who play games will be able to enjoy at the moment. One thing is DLSS, which is still a mundane addition, and another thing is what NVIDIA is trying out with the Path Tracing kit they showed off at GTC that currently only works in Cyberpunk Overdrive.
We should consider its implementation as a mere curiosity for now, no DXP style standard, DirectX Path Tracing or Vulkan extension has been created that anyone can use. In any case, the steps to follow will be the same as those given by NVIDIA, but without binding developers and users to a single brand.
Although having the necessary power, it will still be necessary to wait a few years. In any case, it is very likely that we will see a good part of these functions being applied within the hardware, with new units specialized in said tasks and improved versions of the current ones.
There is no doubt that the implementation of Path Tracing is the next evolutionary step, but there are still years before its massive standardization and not a single generation of graphics cards.
