D3D raytracing not unique to 2080, as Nvidia brings it to GeForce 10, 16

Enlarge / A screenshot of Metro Exodus with raytracing enabled.

Microsoft introduced DirectX raytracing a 12 months in the past, promising to deliver hardware-accelerated raytraced graphics to PC gaming. In August, Nvidia announced its RTX 2080 and 2080Ti, a pair of recent video playing cards with the corporate’s new Turing RTX processors. Along with the common graphics-processing {hardware}, these new chips included two further units of extra cores, one set designed for working machine-learning algorithms and the opposite for computing raytraced graphics. These playing cards have been the primary, and at the moment solely, playing cards to help DirectX Raytracing (DXR).

That is going to alter in April, as Nvidia has introduced that 10-series and 16-series playing cards might be getting some quantity of raytracing help with subsequent month’s driver replace. Particularly, we’re speaking about 10-series playing cards constructed with Pascal chips (that is the 1060 6GB or greater), Titan-branded playing cards with Pascal or Volta chips (the Titan X, XP, and V), and 16-series playing cards with Turing chips (Turing, in distinction to the Turing RTX, lacks the additional cores for raytracing and machine studying).

The GTX 1060 6GB and above should start supporting DXR with next month's Nvidia driver update.
Enlarge / The GTX 1060 6GB and above ought to begin supporting DXR with subsequent month’s Nvidia driver replace.


Unsurprisingly, the efficiency of those playing cards won’t match that of the RTX chips. RTX chips use each their raytracing cores and their machine-learning cores for DXR graphics. To realize an acceptable degree of efficiency, the raytracing simulates comparatively few gentle rays and makes use of machine-learning-based antialiasing to flesh out the raytraced photographs. Absent the devoted {hardware}, DXR on the GTX chips will use 32-bit integer operations on the CUDA cores already used for computation and shader workloads.

Nvidia says that Turing and Pascal playing cards will take two to 3 occasions longer, respectively, to render every body than a Turing RTX card. This distinction is especially acute on Pascal playing cards. In Turing, the 32-bit integer workload used for raytracing can run concurrently with the 32-bit floating-point workload used for different graphical duties. That is not the case on Pascal, the place the workloads must be run consecutively.

This weaker efficiency implies that Nvidia recommends that builders use solely easier raytracing results on the older chips. On the RTX components, the raytracing efficiency might be adequate to allow world illumination—a type of raytracing that permits oblique lighting from reflections, along with the same old direct lighting from gentle sources—however on the GTX components, Nvidia recommends the usage of easier duties corresponding to material-specific reflections.

With the RT cores for raytracing and the tensor cores for machine-learning algorithms, the Turing RTX (bottom graph) can process a frame with raytracing relatively quickly. The Turing (middle graph) doesn't have the dedicated cores, but it can still run the integer workload simultaneously with its floating-point workload for a total frame time of about double the RTX. The Pascal (top graph) has to run the integer and floating-point tasks back to back, so it takes much longer than the Turing, let alone the Turing RTX.
Enlarge / With the RT cores for raytracing and the tensor cores for machine-learning algorithms, the Turing RTX (backside graph) can course of a body with raytracing comparatively rapidly. The Turing (center graph) does not have the devoted cores, however it will possibly nonetheless run the integer workload concurrently with its floating-point workload for a complete body time of about double the RTX. The Pascal (prime graph) has to run the integer and floating-point duties again to again, so it takes for much longer than the Turing, not to mention the Turing RTX.


That raytracing might be carried out on these older chips is not an enormous shock. Throughout DXR’s improvement, Microsoft used a compute-shader-based raytracing algorithm, so clearly the devoted {hardware} is not obligatory. Nevertheless, the substantial efficiency distinction reveals that the devoted {hardware} goes to be vital, a minimum of for now.

Present video games that use DXR ought to mechanically begin utilizing raytracing on the 10-series and 16-series playing cards as soon as the drivers are up to date, with out requiring any sport updates. Nevertheless, given the efficiency differential, we’d think about that builders will need to tailor their raytracing to the older {hardware}. The burden imposed by raytracing on the RTX chips is substantial, and it is merely not going to be sensible to attempt to use the identical degree of raytracing high quality on the older playing cards.

Source link

Show More

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *


Adblock Detected

Please consider supporting us by disabling your ad blocker