Rumor: Radeon RX 6000 Series as fast as a standard RTX 2080 Ti?

This rumor is bad for AMD - even at rumor level - again comparing their new generation product to an older generation product of the competition (they could compare with the "rumored" benchmarks of the new Ampere series since it still at rumor level)...I'm tired of this - "Oh, yes we still can't fight them but look at our prices.. competitive and so on...wait for polaris because this time we will rock, wait for vega because this time we will really rock, wait for, wait for"...

I'm waiting the next rumor abour their ray tracing capabilities - also at RTX2080Ti level???

 

I think nvidia pushed the boat out that far this time it's making all else before it look so bad, surely AMD ain't happy right now, GPU side of course.

 

Custom rdna2 is still rdna2.

 

Nvidia may well have pushed the boat out, but overclocked 2080 TI cards have been available for quite a while now - so existing performance was hardly a secret.

It makes no sense to show a product that falls short, and to assume this is the best AMD have up their sleeve. I suspect AMD knew this 'leak' would be seen and it was deliberately released...

 

I can see a bunch of tech-newbies shitting their pants over non-information. We have no onformation which chip it is, therefore useless rumor and IMO not worth posting.
It's long known the new AMD top chip will be at least 50% over 2080Ti performance and it will be very energy efficient, unlike RTX3xxx series.
Now hold your horses and wait for October.

 

Lol, I remember my first rumor.

 

Coreteks gives a good explanation about the confusion over AMDs performance potential and the SKUs in their lineup.

 

It's not the RDNA part that is custom. Yes, you can order custom SOC solutions from AMD, but you can't customise the architecture (RDNA, ZEN2) itself. Yes, you can specify what performance you want, how musc CU-s, but the baseline achitecture will stay the same and as you specify, it will change accordig to how that specific architecture could scale. For example, you can't get a custom RDNA chip with doubled ALUs inside. That thing is achitecture-locked. You can buy the license instead of ordering the chip if you want, but then you will have to design you chip on your own and bare the consequences.

To be honest, one of the main reasons why Sony and Microsoft is "stuck" with AMD is that no other pary will do what they are doing. Nither Intel nor nVidia accepts this kind of custom orders. You can choose what they're selling or buy the license and design the chip you need on your own. And that is a lot of money you don't want to spend as long as there are at least one company you can go to to do it instead of you.

 

Price will determine who has the better deal, but unfortunately max performance sell mindshare.
I sincerely hope that this is the worst card, the cut down variant, and we can see at least 10 to 20% more from a full working GPU.
I doubt AMD can beat the 3080 in performance, but it needs to be awful close in performance and competitive in price to make a statement.

 

As long as the drivers work well and there are no issues, then that's a start. When 5700XT came out there were a lot of issues and a lot of returned cards comparitively.

If they can have good drivers and are well priced then that is a start. I wonder how TSMC is price wise compared to Samsung ?

 

All I want is 4K @ 60FPS for a reasonable price. Regardless of rumors, AMD does seem to have something that can outperform the 2080Ti, and that I feel is a good (albeit, not great) 4K GPU. The 3070 is a great step in the right direction in affordability, but if AMD can release something competitive for an even lower price, I'll probably get it (just likely not on release day).

AOTS is definitely more of a CPU benchmark, but, a 9900K ought to be fast enough to not be a bottleneck.

Your idea of "keeping it real" sounds like jumping to a lot of conclusions with high confidence, despite not having any real data.

 

RDNA2 is a feature set and not all of those features are present in the PS5 or Xbox, both of which deviate the GPU design in differing ways.

neither console are specifically RDNA or RDNA2, they are rdna-X based and customized to suit the manufacturers whims.

You can say Based On, but you cannot say they ARE RDNA 2.

no.

 

I fairly confident that AMD is going to deliver competitive cards that offer more than 80 or 90% at lower prices, it´s just a question of seing how good and cheap they are against the 3000 series.

 

I just want something in the same performance range of 3080 from AMD at a better price. Doesnt matter a difference of 5-10%

 

How do you know this?

 

Doesn't matter if AMD cards are 2x faster than Nvidia cards, if their drivers are still unstable, horrendous, pieces of sh*t, they're not worth the boxes they arrive in.

 

I personally think the news is fake. Ill just wait until AMD announces and we see actual reviews from Hilbert and others.

 

So many defending AMD like they know exactly what AMD is bringing. As far as I am concerned I want AMD to deliver a 3080 contender. I don't expect that they will though. I think at best we get a card between 3070 and 3080 especially since there is space there. AMD hype on GPU side has been overwhelming while their product has been mostly underwhelming for us that don't want a mid range card. Navi is good for the price, Radeon VII was a joke for gaming yet they still hyped the VRAM (which doesn't do anything for the card in games). They never delivered anything over the 1080 Ti which was a major bummer. 3070 contender is really all AMD needs to make sadly. I'd love to go Zen 3 and RDNA2 but not just to go all AMD.

 

If this is true this is the downfall for amd on the high-end gpu market.
Besides that i still get a ps5 tho its cheap and i loved my ps4 pro in general.

 

In my case, that's a non-issue. AMD's Linux drivers tend to actually be overall better than Nvidia's (not so much at release date but not months later either). I don't understand how AMD screws up their Windows drivers so badly though. Their products release rather late no matter what, so why not just delay another month to make sure the drivers aren't pissing people off? Though, maybe that's what they're planning with RDNA2. The good news is, it's highly unlikely the drivers will be worse than RDNA1. It's not an all-new architecture so it's really just a matter of compensating for the differences and fine-tuning everything else.

I'm sure the results are real, but they're not meaningful because we don't know what model(s) are used and AOTS wouldn't exactly be one of my top picks for testing GPU performance. Too many unknowns to draw a conclusion, good or bad.

Huh? AMD hasn't been competitive in the high-end market since 2014. The downfall has long passed, to the point that they didn't even release a "real" high-end product for the 5000 series.