By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Cyran on 01 June 2021
Edit Thread
hinch said:
Cyran said:

What would be the point of bringing something like that when most games already got something like that as a option.  it would get 0 traction.  They needed something that they could at least say they got a alterative to DLSS since DLSS been getting so much praise lastly.  Also maybe FSR is a good building block to take in account motion vectors in the future which those other solutions would not of been.

All am saying is my experience is 9 out of 10 times when the initial reaction is that don't look good on paper and I hear just wait too see it in x or y condition because surly such a large company know what they doing it turn out nope it exactly what the first impression was.

I hope this is the 1 out of 10 times that I am surprise and it actually turn out it better then what it appears to be but I wont believe it till I see it.

The point would be capturing the marketshare and mind share since AMD did not have any comparable technologies in their products. This will be AMD's first dive into this and they have to start somewhere. I mean there is a reason why they shown the GTX 1060 in their presentation since that card is most widely used (if we go by Steam survey) by PC gamers.

The bigger the userbase the bigger the audience. FSR will be optimised for AMD hardware no doubt and with RDNA 3 incoming, its a good time roll it out. Who knows maybe big changes and improvements for upcoming cards. Not saying it will but yeah I find it quite exciting stuff. For closed based systems like consoles and laptops/portable especially.

Though yeah, it does look quite rough from the screenshot

We on the same page.  My what the point comment was directed at haxxiy about why AMD would not use checkerboard rendering and temporal upsampling if FSR was not any better out of the gate.  My point was FSR could be the building point in future for something as good as DLSS and it give them something they can claim as a competitor to DLSS IE get some market share as you put it even if it don't compete very well out of the gate. 

I was making the case there no point for them to release a checkerboard rendering and temporal upsampling but there FSR solution make a lot of sense even if it not very good on first release.  For the reasons you stated.