Tested: Nvidia’s Variable Rate Supersampling Doesn’t Deliver on its Promises (Yet)

I’m glad someone took the time to test the feature but the author is extremely confused about what VRSS does.

It improves IQ vs full screen MSAA with reduced performance. It increases performance vs full screen SSAA but at reduced IQ.

It is not expected to increase performance vs full screen MSAA. If the foveated area is super sampled and the rest of the frame is multi sampled obviously performance will be lower than pure multi sampling.

The article also claims VRSS uses tensor cores and that is clearly false.
 
Last edited:
fightingfi said:
https://www.tomshardware.com/features/nvidia-vrss-vr-testing

good reading
Click to expand...

Apparently it was such a bad read they took it down:

upload_2020-1-27_20-22-15.png
 
I saved it as a pdf
 

Attachments

  • Tom's NV VRRS.pdf
    1.9 MB · Views: 0
Enigma said:
God damn, This is funny... I miss [H]ard|OCP, Kyle and the boys always gave a fair unbiased review..... I miss the days of yore..... FrgMSTR Where are you when we need you?
Click to expand...

Yeah....we even got posters on this forum disliking facts from reviewers...because the feel their own opinion is more valid...people are getting more and more retarded by the day.
But I do look forward to seeing take two from Tom's on this....will they acknowledge the manjor brainfart...or try and sweep it under the floor?
 
Factum said:
Yeah....we even got posters on this forum disliking facts from reviewers...because the feel their own opinion is more valid...people are getting more and more retarded by the day.
But I do look forward to seeing take two from Tom's on this....will they acknowledge the manjor brainfart...or try and sweep it under the floor?
Click to expand...

Even if they do fix or even acknowledge their failure, it doesn't take away from the stream of halfassed tripe that flows onto their site.


And that's not accusing them of not meaning well, just an observation that there appears to be a very large experience gap in their review staff. I get that content drives clicks, but proper review methods take both practice and experience to get right.
 
IdiotInCharge said:
Even if they do fix or even acknowledge their failure, it doesn't take away from the stream of halfassed tripe that flows onto their site.


And that's not accusing them of not meaning well, just an observation that there appears to be a very large experience gap in their review staff. I get that content drives clicks, but proper review methods take both practice and experience to get right.
Click to expand...

I never really liked Tom's...first came across them in the P4 Northwood age...and I wasn't impressed back then...remember their forums being quirky and poorly designed...I hope they have gotten better, but this doesn't really indicate it.
 
Factum said:
I never really liked Tom's...first came across them in the P4 Northwood age...and I wasn't impressed back then...remember their forums being quirky and poorly designed...I hope they have gotten better, but this doesn't really indicate it.
Click to expand...

Their 'height' was really before that point. I'd already written them off and preferred [H] balanced with Anandtech (now mostly focused elsewhere) and TechReport (not going back after their re-org).
 
Yeah Toms was the hotness back in the GeForce 2/3 days. They hit rock bottom a few years ago and are better now but nowhere near their peak.
 
You must log in or register to reply here.
Back
Top