NVIDIA and AMD Called Out for Strange Graphics Card Benchmarks

[Tsing]

Image: NVIDIA



Enthusiasts and tech publications are calling NVIDIA and AMD out for publishing weird benchmarks that seem either unnecessary or, worse, make their newer products seem underwhelming compared to their predecessors.



NVIDIA, for one, decided to update its landing page for the GeForce RTX 3050 with benchmarks that pit the new budget option against the GeForce GTX 1650 and GeForce GTX 1050, two older graphics cards that were initially released in April 2019 and October 2016, respectively. Naturally, the benchmarks confirm that the new GeForce RTX 3050 blows away its older siblings, particularly when ray tracing is thrown into the mix...

Continue reading...

 

[Zarathustra]

I mean, Nvidia is just doing what Nvidia always does. Try to manipulate things to make them look better.

Remember when the RTX 2xxx series was launched and they were threatening to drop review sites that talked about raster performance?

AMD - however - needs to fire their entire marketing department. It's like they have lost their collective minds.

In addition to the Radeon Pro comparison, look at the RX6500xt benchmarks, and how it performs WORSE than the previous gen rx 5500xt.

They clearly should have done something different here. Maybe name it the rx 6400xt?

It's a decent budget card that is getting absolutely torn apart in gaming circles right now because "the current gen is worse than the previous gen", when that was really just a naming choice.

AMD's marketing department seems to almost be trying to sabotage themselves.

 

[LazyGamer]

I mean, Nvidia is just doing what Nvidia always does. Try to manipulate things to make them look better.

That's really just marketing. I get that people feel slighted when it happens with products they have an interest in, but it's good to keep in mind that these are businesses, publicly held, that are beholden to their owners - their shareholders.

Realistically, we should just plain expect these 'distortions'. They should still be called out - no free passes! - but if the data being presented wasn't the result of truly independent testing, then it should not be taken literally.

But let's take that RTX 3050 - I look at the benchmark provided and don't see anything particularly suspect. Sure, it has RT and its 'predecessors' didn't, but it also has DLSS. The RTX 3050 is the card that needs DLSS the most!

Remember when the RTX 2xxx series was launched and they were threatening to drop review sites that talked about raster performance?

Scummy, and really pretty boneheaded on Nvidia's part. Yeah, they had to figure out how to push RT into the market on their own, and it was definitely a 'new' thing that reviewers really didn't want to dwell on. Mostly because everyone including Nvidia knew that it was immature.

AMD's marketing department seems to almost be trying to sabotage themselves.

It's hard to remember when this wasn't the case - though I don't envy their jobs.