Nvidia GeForce RTX 5070 review: No, it’s not “4090 performance at $549”

It's not all bad news, but in a lot of ways this is barely an upgrade.

See full article...

 

[David Mayer]

Pedantry against the trend: "interpolated frames", these are extrapolated frames, interpolation draws from the data on both sides of the created data point (the frame or frames before and the frame or frames after), extrapolation draws from the data on only one side of the data point (the frame or frames before or after) not both. This tech extrapolates as it completes generation of the new frame(s) before subsequent ones exist.

All that said, I'm late to the party and there's no hope of overriding this linguistic trend, so, sucks to be a pedant I guess.

Edit, I'm wrong, and nvidia is actually interpolating, according to this video at 12:58 they are analyzing two frames and generating frames in between them. I made the mistake because I assumed they wouldn't just throw away latency to get these results.

there's really you know two uh rendered frames that that were analyzing in order to create a series of of frames in between

 

[David Mayer]

This tech (DLSS 4 MFG) does NOT extrapolate, it still very much interpolates as mentioned by Nvidia's own VP of Applied Deep Learning Research Brian Catanzaro :

View: https://youtu.be/uyxXRXDtcPA?t=768



There are plenty of deplorable linguistic trends out there, this is not one of them.

Can you please link the relevant timestamp, I'm not going through a 30 minute video for this.
Edit1: Took a quick skim, from your video, a graphic showing extrapolation:



Edit2: Why TF would you interpolate anyway, that would deliberately add latency.

Edit3: This was driving me a little nuts, I found the timestamp, 12:48, thank goodness for searchable transcripts.

there's really you know two uh rendered frames that that were analyzing in order to create a series of of frames in between

In the future please include a timestamp and a quote in your video citations.

 

[David Mayer]

The time stamp is 12:47, the marketing is intentionally deceptive. Jensen claims on stage that it can predict the future.

Thanks, this is really quite absurd, the tech to "predict the future" actually does exist and is used now for VR. Why TF are they deliberately adding latency?

 

[David Mayer]

While I'm here, does AMDs product do the same thing?

 

[David Mayer]

Thanks for the kind advice.
In the future before ranting about linguistic trends please make sure you're factually correct.

So don't trust communications directly from the source, got it.

 

[David Mayer]

I guess for any playback that isn’t responsive to user input. Seems like a post-processing interpolation tech they got to scale to the point of working in ‘real-time’ that turned into a marketing point.

Your brain rebels against the concept because putting it front and center is scummy and moronic.

I guess for any playback that isn’t responsive to user input.

This tech is specifically markets as a tool to enhance interactive gaming.

 

[David Mayer]

You've realized that it is actually interpolation and the next step to realize what impact that has on latency. because it has to wait for 2 native frames before giving 4 frames to the user, it has to add 2 native frames of latency to the output.

What makes you think I haven't realized that?