An article published by Igor's Lab confirms that NVIDIA's Frame Generation technology is applied at the tail end of the frame processing pipeline and perfectly compatible with FSR 2.0 and Intel XeSS AI Up-scaling pipelines. While there isn't any real advantage to doing this, it didn't stop Igor from taking some pioneering benchmarks of what the numbers would be like and surprisingly, it seems that FSR + Frame Generation can actually beat DLSS + Frame Generation in pure performance figures.
AMD FSR + NVIDIA Frame Generation beats DLSS in NVIDIA RTX 4090 benchmarks
Let's begin with some context. Frame Generation is a technology that is part of NVIDIA's DLSS 3.0 stack which generates artificially created frames between two frames generated by DLSS' AI implementation. Think of these frames like padding between two computationally generated frames. As it turns out, if you are an NVIDIA 4000 series GPU owner, you can actually use NVIDIA frame generation with frames that re being computationally generated by Intel XeSS or AMD FSR 2.0.
The picture quality between AMD FSR 2.0 + NVIDIA Frame Generation and DLSS + Frame Generation seems fairly comparable although NVIDIA Frame Generation and Intel XeSS appear to be on the softer side of things. So without any further ado, here are the benchmarks from all three configurations:
Keep in mind fps numbers only tell part of the story, to get a full idea, we would encourage you to head over to Igor's Lab and read up on frame times, variations and percentile figures of the three systems but for now, lets focus on the maximum fps achievable. With DLSS and FG (Frame Generation), you are looking at a maximum value of 224 fps. Pretty decent right? Keep in mind even this value may be seriously software bottle necked due to frame time limitations in the game engine.
Up next, we have AMD FSR chugging out the frames in the back end and NVIDIA Frame Generation being used to fill in the gaps. The maximum achievable FPS here in Spider-Man Remastered is 231.4 fps, which is slightly higher than the native approach of using DLSS. That said, this does seem to be a software bottleneck situation because other metrics are actually fairly comparable to the DLSS based implementation.
Finally we have the Intel XeSS + NVIDIA Frame Generation based run which yields a maximum achievable fps of 203.9 - which does not seem to be a software bottleneck. As noted above, Intel XeSS is also slightly softer in the picture quality department than either FSR 2.0 or NVIDIA DLSS.
One thing is for sure though, further testing in titles that support this will be required to check if FSR might actually increase performance when coupled with NVIDIA Frame Generation on RTX 4000 series GPUs - in a title which does not run into software bottlenecks. Cross vendor compatibility between up scaling technologies is something pretty exciting and could result in a best of both worlds outcome for consumers.
