Nvidia giving free GPU samples to reviewers that follow procedure

Of course, it hasn't been reviewed before, because it was not a challenge for even an IGP. Useless for GPU benchmarking.
However, Minecraft RTX changed everything as the first fully path traced game that was not a demo developed by Nvidia (ie Quake RTX). Being so popular and technically advanced, it's the perfect case to measure RT performance across different GPUs

 

Since games began being tested with beta implementations of DX12. The only reason Minecraft RTX is listed as beta was because of ongoing DXR support, but in it's heart it is just the full game with DXR.

 

Raytracing in Dirt 5 was only available with a non public beta branch. Wasnt a problem for HBU...

 

First: do ease down on the accusations of hysteria please.
Second: you speak as if reviews hadn't been out there for everyone to see?


I don't even need to pick on 4K. The >$1500 RTX3090 gets 30 FPS at Ultra-Wide QHD:






And the 2080 Ti that cost >$1200 up until a couple of months ago gets 24 FPS at 1440p:

 

In which review? Or do you mean using press build of a game which is common practice in near-launch situations to give reviewers time with the game

 

Bad move from nvidia. Hardware unboxed is well know, and that kind of thing won't fly with the hardcore pc types that buy the high end stuff. I understand where Nvidia is coming from. If you're sending thousands of dollars worth of gpus to reviewers and they fail to adequately review the entirety of what it offers, you probably won't see the purpose in continuing the relationship. Just the optics are bad. Hardware Unboxed really should be thoroughly testing ray tracing, whether they like the feature or feel it's worth it. It's a major feature of both Ampere and RDNA2. It's part of what you pay for when you drop these high prices. It doesn't really make sense to half ass it just because your viewers currently don't seem to be interested. I like the site and their content.

 

Which they are.






This video released in the same day they got the boot from nvidia, so it's not like they're reacting out of being blacklisted.

 

Its horrible that big companies like nvidia withold product from people who dont give favorable reviews.

Additionally, ignoring raytracing performance in 2020 is like ignoring programmable shaders in 2003, completely goofy. They should be a laughing stock, not blacklisted by nvidia. Cannot believe how much fans have managed to downplay RT.

 

HBU used Dirt 5 in the 6800XT review:

It isnt "common practice" to use non available beta branches of games sponsored by the competition and ommiting open beta software from the other company.

 

60fps maxed native 4k using DLSS. Wouldnt call that terrible RT performance.

DF disagrees, they called it the new Crysis.

 

Erm.. Dirt 5 was released with RT support 2 weeks before that video. How is it "non-available beta branch"?

There is no such thing. With DLSS you're not rendering it at 4K

 

This is a low move by Nvidia, and Gamers Nexus nailed it:

This is such a bad and dishonest take.

I'm sorry, but someone needs to say this: you are one of the biggest Nvidia apologists on this forum, and it makes it very jarring to read discussions on Nvidia [and AMD] technologies whenever you're involved. This isn't an ad hominem -- we all have our our biases and soft spots -- but it's something to be aware of and maybe think about before getting into a discussion about TWIMTBP, SLI, GameWorks, DLSS, or RT.

Way to take their quote out of context. HUB specified that this was because SAM (and the Nvidia equivalent, which they also praised) applies performance improvements across all titles. Performance with SAM can only be greater than or equal to performance without it, with no impact to image quality or frame time delivery.

 

Dirt 5 was released. Ray tracing is still MiA.

 

Yep it's a slide show fully maxed out. As with all games there are probably things you can turn down for no noticeable IQ loss. Really looking forward to the DF review, so far the coverage has been pretty superficial.

I don't think this is completely true. Several reviews have shown performance loss with SAM enabled.

 

@ToTTenTranz I watch all of their reviews. I find them useful. Ray tracing really isn't a major part of their reviewers. They do cover it, but it's pretty sparse. The 3070 review gave two minutes to both ray tracing and dlss, and only showed performance in 1 ray tracing title, Shadow of the Tomb Raider, which is known to not be an interesting implementation of the feature. Most of that two minutes goes to DLSS performance in Death Stranding and F1. I get it, they don't care much about ray tracing, and they don't think their viewers do. But when you're getting things free for review, it's really a two way exchange. Hardware unboxed is not entitled to free products for review. Nvidia is giving them out for something in return. Nvidia wants certain features to be tested in exchange for providing free product. Hardware unboxed could spend more time testing it and in the end say, "We tested it extensively and we don't think the performance loss is worth it." In that case if Nvidia stopped giving them free cards because they didn't like the conclusion I'd say nvidia is fully in the wrong. What we have right now is just a difference of opinion between nvidia and hardware unboxed about what needs to be tested in exchange for free products.

 

Confirmation bias in action I would say, indeed.

 

Hello, first post so go easy

• Nvidia in this instance isn't claiming misrepresentation, but insufficient representation in its favour. It believes it should have direction on the methodology of the reviewer, i.e. the selection of games. I am surprised that people on such a reputable site would be willing to entertain that interference. Not all reviewers used SAM remember
  
• The specific request here obviously ties into the fact that rasterization parity with a competitor needs to be offset. I'd be surprised if members of this site demanded that AMD be allowed to force no RT titles in its graphics cards reviews to enhance its rasterization results. You're willing to swing one way only?
• If Nvidia hasn't drummed up enough developer support for RT, is it the responsibility of the reviewer to maximize RTX representation to ensure Nvidia is sufficiently covered?
• If RTX is the new standard why should the overwhelmingly rasterized based library of existing games be ignored?
• DF is being touted as a brilliant source. Yet they opted to for AC: Unity and Odyssey instead of Valhalla. Why is that?

Thanks.

 

Dunno, NV still has the rasterization advantage before counting in DLSS and RT.

 

@Zoal I get it. Nvidia is being a dick because they know they lead in ray tracing performance so they want it shown favourably in reviews. They want reviews that look at ray tracing extensively because they see it as a major feature and selling point of their cards.

On the flip side, Hardware Unboxed barely looked at ray tracing in their last two Nvidia FE reviews. The 3070 review showed performance in Shadow of the Tomb Raider. The 3060ti review showed performance in Watch Dogs Legion. If a reviewer was testing rasterizer performance and they did it by showing you benchmarks for one game, you'd say the review sucks. Hardware Unboxed's ray tracing coverage for the last two Nvidia releases was bad for the consumer, not just nvidia. I know they're managing their workload, and they're trying to keep the reviews short, but I can see why Nvidia would not want to give them free cards. I still think Nvidia should, because hardware unboxed is a relatively big review channel now.

 

Should AMD reserve the right then to push games with superior optimization for its ray-tracing solution? What is the acceptable level of corporate demands on reviewers acceptable to you? What if I conjure scenarios to point out weaknesses in Nvidia's ray-tracing or DLSS solution. Should that be considered inappropriate representation and grounds for dismissal?