Nvidia giving free GPU samples to reviewers that follow procedure

HWU says the value of DLSS is questionable because it may not ever get wide spread support. Not because they don't find the improved 2.0 implementation to be impressive. Steve also says it’s just his personal opinion that RT currently isn't that big of a selling point given what it offers in the currently available games. He finds Legion to offer a good visual uplift but most other games to not be worth the huge performance hit. It’s not a far fetched viewpoint. Most of what we have seen prior to Legion has not been particularly transformative, especially when considering the performance sacrifices.

 

If you are at this moment arguing whether a reviewing webside does or does not have merit on their opinion about RT + DLSS, then Nvidia reached their goal, and you were its pawn.

They aim their marketing decisions (this was a marketing decision, no doubt) to the people who let deeper meanings go above their heads.

From what I'm reading on these forums, nvidia has accomplished their goal. Linus nailed it.

 

Well then, from now on DF and other reviewers should totally ignore all RT in games, in special on console, since the offering in favour of performance isnt worth it?

 

Steve did not see through the real intentions of that e-mail. Instead of and keeping the situation quiet, he played the game exactly like Nvidia intended.

They are now having a boatload of marketing for free, while throwing a warning bullet to every other publication.

What premeditated this situation was AMD excepcional recovery, which prompted Nvidia to act like te old days. AMD needs to star winning real soon for everybody's sake.

 

Stating the obvious is wrong now? How is it ignorant?

whooosh

 

Lol well damn.

 

And that is the problem. Control costs 20€. Has support for DLSS and Raytracing. He is basing his conclusion on his own bias and ignoring the fact that buying a 6800XT over 3080 means that you have a worse experience in a numbers of games. HBU even used Control in the 6800XT review. So, why did he not used Raytracing and DLSS in a game he is using to compare nVidia with AMD? Does it make even sense to use games with Raytracing and DLSS support in a benchmark parcour and not using all options?

 

He mentions Control as the other option where RT has merit. Buying a 3080 over a 6800xt means you have a worse experience in a number of games. Will there be people who play Control without RT? I'd assume yes so it makes sense to have results without them. Its obvious why you wouldn't compare DLSS results to native. There are no problems comparing results with RT on btw, I'm not arguing against that. His personal opinion on Control is that he would play with RT turned off and DLSS on to achieve high framerates FYI.

 

Why does it matter how consumers play their games? Console games have different settings now, too.
Control supports Raytracing and DLSS. Control is used in a comparision between a 3080 and 6800XT. But the reviewer disabled features because he thinks they are not worth it.

You see the problem? Personal bias dictates how he is comparing different hardware.

 

A wild guess is that he has disabled RT if possible in all his PS5 games.

 

I don't feel tech media primarily (well for the most part) really caters primarily to the end consumer. Yes to some extent it does guide purchases for some people but would you really say the majority of content consumption is from that angle? I know personally I consume considerably more review content than products I'd even consider purchasing. I'd wager reviews are used more in online "discussions" by "spirited" commenters more than guiding their own purchases as well.

Despite the fact that this action by Nvidia on the surface is a "negative" against HWU, or tech media in general, on the whole it's likely beneficial. "Content" like this is really what drives higher engagement and viewer/readership. Cold data and neutrality itself is boring for most people.

 

It’s fine for reviewers to share subjective opinions. That’s what we expect them to do. The problem with HWUB is that they seemed to go out of their way to paint RT in the worst possible light by using the crappiest RT games in their suite all while giving air time to mediocre titles like Godfall. They’ve since addressed that by using games that actually have decent RT implementations. Now they can continue that trend and add Cyberpunk to the mix.

Even after all that it’s still ok for them to think that RT isn’t worth it. Other reviewers including GN have also been less than enthusiastic about RT but Nvidia isn’t giving them shit for some reason.

It will be interesting to see HWUB’s take on whether RT in Cyberpunk is worth it or not.

Interesting point. I would think that the YouTube crowd (Linus, Jayz, DF) reach a far more diverse population than the written articles. I only really noticed them a few months ago and was amazed at the following they have.

So yes I agree that review content is often used as fodder for online debate but I also think a lot of people rely on those reviews to make purchase decisions. E.g. in my case I was definitely swayed by the reporting on the quality of ASUS Ampere cards vs the other AIBs.

 

They think RT is good and worthwhile but not a game changer like some have said. They think DLSS quality is the best option as its similar to native. Better in some areas and worse in others.

 

As long as the settings are the same for everyone in the test, there is no problem whatsoever, why should there be?
Of course it would be nice to get results with and without RT, but as long as the reviewer keeps the settings the same for everyone tested, it's a fair game. If you don't appreciate the settings reviewer x wants to use, don't read their reviews, but don't act like there would be something wrong with it.

Using DLSS in comparison graphs between AMD and NVIDIA would be a problem, it's a point to be brought up separately that you can trade image quality for performance (and you might even like that image more depending on your preferences)

 

Very nice to know where people stand, all false modesty and objectivity thrown off now I wonder if the same people who criticize Hardware Unboxed for their opinions and test methodology did do the same when Deep Foundry did their "200% in Doom" video or when some media outlets "reviewed" 3090 in 8K on a sponsored $20k 8K OLED TV (it's like 2x the median yearly salary here, try to refuse such a gift, considering this "editorialize as we like or else" appoach to reviewer relations by the global marketing head honcho at nV)

 

HWU also did an entire video dedicated to RT and DLSS.

 

Gameplay wise, maybe. But otherwise graphically its up there. Its really amazing in special at some scenes.

RT is def a gamechanger in this one, in special indoors. DF and Linus thinkso, atleast.

 

That's not what I called absurd.
Reread my previous response, and stop trying to construct a straw man.

Gross.

Gross, again.

You can't be serious. Surely you were intoxicated when you posted this.











https://www.youtube.com/watch?v=YH5QNQngL8A

https://www.youtube.com/watch?v=NxvoxpeeIaQ

https://twitter.com/hardwareunboxed/status/1337699116361658368?s=21

Yeah man, they totally ignore ray tracing.

Again, there are parallels between the hype surrounding ray tracing today and something like DX10 or tessellation back in the G80 vs R600 days.

Let's be clear: no one is denying that RT (and ML-based SS like DLSS) is the future of games. The issue is that, so far, most implementations of RT are not worth the drastic hit to performance for a negligible improvement in IQ. As devs learn to better utilise RT (thanks to Nvidia's RTX, the consoles, and Navi 2nd gen), this will change. I think games like Control and CP2077 are previews of what's to come in the future.

The DX9 -> DX10 move was huge. The move away from pixel shaders and fixed-function units towards a more generalist approach to GPU processing was a massive game changer. G80's performance was incredible and RV770 was amazing. But those first couple generations of DX10 and DX10.1 implementations sucked. But here we are, finally, with DX12 and Vulkan.

RT will follow the same path.

The sense of entitlement of Nvidia and its grifters regarding tech reviewers is both funny and a little depressing.

 

So, you are saying a reviewer can ignore raytracing and should be entitled to get free samples from a company who is pushing raytracing? Sounds like a paradoxon.

No, using DLSS is fair. A reviewer should use a "performance" equal resolution on AMD's side so that viewer and reader can compare image quality. I mean when raytracing is bad because of the performance hit, rasterizing in higher resolution doesnt make any sense, too.

 

Yeah I thought it was a good video. They acknowledged the trade offs of DLSS vs native and basically recommended DLSS Quality for all Nvidia users.

Noticeably they didn’t comment on whether RT or DLSS should influence your purchase decision of Nvidia vs AMD.