No DX12 Software is Suitable for Benchmarking *spawn*

Even in the orginal game the content between a 2060 and 5700XT is not the same: https://imgsli.com/NTk1OTY
And the EE improved a lot with the fully raytraced infinity bounce GI system.

My remark with DX12 is your argument. I never have seen huge difference in the content between a DX11 and DX12 path. So i guess DX12 is useless and should never be used in benchmarks?

 

This has nothing to do with Nvidia. RDNA2 benefits from the same experience that RTX cards have.

RDNA1 is just a horrible, outdated architecture and it's going to show more and more as we tap into the potential of this generation. Metro Exodus Enhanced Edition is just a taste of what's to come...

 

Enough with the platform warring. Move on to something else. Those who have demonstrated interest only in platform warring shall be banned.

 

I don't see the point in having software support for a feature that would be considered too slow to be usable, no. You might as well start demanding every single hardware feature in future DX/Vulkan to be supporter on old gpus via compute.

I haven't seen Pascal users being upset that their GPUs typically aren't featured in the new raytracing benchmarks either.

It's natural that a new game not using dxr is included whereas another game requiring it is left out on a gpu that cant run it. IF anything, the version they would havet tested is the original version.

 

It's not pointless, because next gen games might actually use DXR as their main lighting solution. That Avatar game is already confirmed to do that, so it has to run in software on GPUs like RDNA1 and Pascal. I do wonder if AMD will enable DXR for the older cards eventually, otherwise these cards might not be able to boot up future games at all.

 

The diskussion of Metro Exodus Enhanced is why I brought up previous examples from past, users didn't say the GTX 260 wasnt able to run games anymore when it couldn't run eg Crysis 2's separate dx11 exe that arrived later, they were considered relevant as long as they could run the dx9 or dx10 exe which they could for years

 

To play next gen games on RDNA1.

 

Or that a 5700xt can't run Doom Eternal or any other game with RT options.

 

I don't see any different content.

 

In over a dozen AAA titles the 5700XT can't achieve even current console level image quality, while the 2060S can, the 2060S can access higher levels of IQ not available to the 5700XT.

A fact that proved nothing on the actual grounds in their opinion, they used 8 games supporting DLSS2 in their review, yet they didn't even care to use DLSS in a single one of them, despite the 2060S being able to immensely benefit from it, especially at 4K resolution.

 

Whoa, so your previous post stated that the 5700XT is locked out from playing 35 titles, and now your further reflection around the subject clarifies that the barrier is set by not achieving console quality, as well as providing a clear answer that accessing higher levels of IQ is more important for every single PC/console user than turning off individual settings for higher performance. You are free to think that, but that hard set opinion that you¨re trying to force onto everyone is why this discussion is going nowhere. people are free to disagree but that doesn't mean you shall invalidate everyone not agreeing with you.

If you don´t like Hardwareunboxed, just ignore it and stick to other content. Their content is valuable for some, and obviously useless to others. That´s not worth wasting further threads on.

 

Some people here often argue that Ultra settings on PC affect image quality in a very negligible way, while costing 30% or more of performance, that's because Ultra settings usually just extend draw distance, resolution of shadows, volumetrics .. etc, HardwareUnboxed shares this opinion too! Well, Ray Tracing is here to change this completely then, RT reflections and GI while cost a lot of performance, really do affect the image quality in a big and SIGNIFICANT way, much much more than any traditional Ultra settings.

If you want to have the best of both worlds, you can easily run just High settings + Ray Tracing and achieve a very solid performance and IQ uplift.

Can the 5700XT run Doom Eternal with the upgraded RT reflections? Or Watch Dogs Legions with it's massively important RT reflections? Can it render the GI in Metro Exodus Enhanced Edition and Resident Evil Village? The increased shadow coverage in Call Of Duty Cold War or The Medium? The answer is nope, but consoles can do all of that.

But HardwareUnboxed would rather ignore all of that, and ignore the added performance from DLSS2, and just casually present the 2060S as an inferior product to the 5700XT even in 2021, despite the 2060S delivering both a superior IQ that is either better or equal to consoles and the path to better performance through DLSS.

LET THAT SINK IN: in 2021 a so called reviewer is calling a piece of hardware that can't even rival consoles a better buy.

 

Yet another post presenting your own opinions as established facts representative of what everyone else thinks.

Not all games do support DLSS, and even for those that do, iťs not guaranteed at launch. You speak of Doom Eternal, it took more than a year until it got DXR and DLSS, a long wait for people who required DXR and DLSS at launch.

Why is it so bloody hard to accept that a review can focus on aspects that differ from the ones you find important and instead have this obsession of invalidating everyone not agreeing with you? No one can force you to like it but this constant claim that iťs wrong and irrelevant for for everyone, thaťs weird.
5700xt can be better for some users, the 2060 for others, it depends on how the individuals in question consider the possible trade-offs and thats what you ought to let sink in.

 

And that's fine, you don't have to show everything every time.

 

F1 2021 Performance At 1080p, 1440p, Ultrawide & 4K – Techgage
July 21, 2021

 

Since 3dMark was mentioned my mboard bios has this option
3DMark01 Enhancement Allows you to determine whether to enhance some legacy benchmark performance. (Default: Disabled)

Anyone have any idea what it does?
edit: according to a youtube video I found it lowers the score by 41. The p.c scored about 8500 at 4k with it disabled