Question Speculation: RDNA2 + CDNA Architectures thread

[uzzi38]

All die sizes are within 5mm^2. The poster here has been right on some things in the past afaik, and to his credit was the first to saying 505mm^2 for Navi21, which other people have backed up. Even still though, take the following with a pich of salt.

Navi21 - 505mm^2

Navi22 - 340mm^2

Navi23 - 240mm^2

Source is the following post: https://www.ptt.cc/bbs/PC_Shopping/M.1588075782.A.C1E.html

 

[kurosaki]

I don't really understand your reasoning. It's clearly better that reducing the resolution manually. Turning down other quality settings cause higher drop in image quality.

I'd rather play in 1080p native than "4k"DLSS, any day of the weak.
Spit the blue pill out while there's still hope left!

AMD Radeon RX 6800 och RX 6800 XT

AMD:s efterlängtade "Big Navi" är äntligen här! På plats i testlabbet är Radeon RX 6800 och RX 6800 XT. Läs hela artikeln här (/artikel/30790-amd-radeon-rx-6...

www.sweclockers.com

 

[kurosaki]

Just because a product is available, doesn't mean it is mainstream. Mainstream represents the largest install base. And that is far and away 1080P, by a gigantic margin. For a lot of people $300 on a monitor is too much, when they can get a similar sized one for $120. They don't know the difference between 1440 and 1080.

And as long as games keep getting more demanding, GPU's will also have to get more demanding. Technically we had 4K gaming cards five years ago. And they will run games from five years ago. But definitely not todays games. So it should not be a surprise that you need an upper end card to play on a higher resolution display. A high end card may be twice as fast as a low end card, but a high end display (4K) has quadruple the pixels of a mainstream display (1080).

I know. I think I'm just disappointed that games always are five years ahead of graphics cards, it gets worse, the more powerful setup you have. With a small 1080p screen, i would probably manage five years with relative cheap cards. If I just by a notch go up to 1440p, the demand on the card grows, and the price gets more expensive for the same settings otherwise, and on top of that, the card grows older exponentially. 4k, with the pricing of cards today, and having the urge to buy next gen before the current has arrived on the mail. No thank you.

 

[Kenmitch]

Not too shabby.

Air-Cooled AMD RX 6800 XT Claims World Record On HWBOT

All hail the new king!

www.tomshardware.com

 

[kurosaki]

When is anandtech done with the GPU deep dive and review anyways?

 

[jpiniero]

When is anandtech done with the GPU deep dive and review anyways?

Right after they finish the 960 review.

 

[zinfamous]

Right after they finish the 960 review.

This expression is now AT's very own "But will it run Crisis?"

 

[kurosaki]

This expression is now AT's very own "But will it run Crisis?"

Cut me all you want!

 

[PhoBoChai]

AMD and Nvidia implementations differ quite a bit, so if someone was expecting RTX optimized games (or tech demos) to run well on 6800 series they are seriously naive. Not saying that Dirt5 is representative of what we can expect, but as in all things thuth is in the middle. And btw, expect ray tracing being the "next optimization war".

Why did ppl even expect AMD GPUs to run RTX + NV optimized games well in the first place?

Have they not learnt anything in the past decade?

NV has shown they go out of their way to gimp AMD GPUs whenever they sponsor a game, and even game engines.

I don't expect this to change or suddenly Jensen to be a nice guy.

 

[Konan]

Thought this view was really interesting -
Performance of the RX 6800 XT relative to the RTX 3080 at 4K resolution (no SMA, Rage mode, RT, DLSS) across 10 media outlets with a total of 120 results (see below for sources)

https://twitter.com/x/status/1329428731933433861

 

[Elfear]

Thought this view was really interesting -
Performance of the RX 6800 XT relative to the RTX 3080 at 4K resolution (no SMA, Rage mode, RT, DLSS) across 10 media outlets with a total of 120 results (see below for sources)

https://twitter.com/x/status/1329428731933433861

View attachment 34184

Thanks for posting. Looks like the average across those 10 sites is a 4% lead by the 3080 at 4k. I'd be interested in seeing the same comparison at 1440p.

 

[Mopetar]

It's also interesting to look at all of the different results like that in one place. It really makes a few things jump out.

Tom's found the 6800 XT to be 89% of the 3080 in Strange Brigade, but TPU found that it was 107% which is a pretty big difference. I'm kind of curious what's going on that we see such a different result. Before someone says "'cause TPU!" there are plenty of their entries that are inline with other publisher's results.

 

[Hans Gruber]

It's also interesting to look at all of the different results like that in one place. It really makes a few things jump out.

Tom's found the 6800 XT to be 89% of the 3080 in Strange Brigade, but TPU found that it was 107% which is a pretty big difference. I'm kind of curious what's going on that we see such a different result. Before someone says "'cause TPU!" there are plenty of their entries that are inline with other publisher's results.

What CPU were they using. What ram were they using and the timings. These are the mysteries of the unknown. Probably a lot of different setups which yield different results.

 

[kurosaki]

It's also interesting to look at all of the different results like that in one place. It really makes a few things jump out.

Tom's found the 6800 XT to be 89% of the 3080 in Strange Brigade, but TPU found that it was 107% which is a pretty big difference. I'm kind of curious what's going on that we see such a different result. Before someone says "'cause TPU!" there are plenty of their entries that are inline with other publisher's results.

DX12 or vulcan differ hugely in some titles. Might be different APIs?

 

[Mopetar]

What CPU were they using. What ram were they using and the timings. These are the mysteries of the unknown. Probably a lot of different setups which yield different results.

Tom's was using both a 5900X and 10900K with 4x8 DDR-4000 memory for the AMD CPU and DDR-3600 for the Intel. TPU used a 9900K with 2x8 DDR-4000 memory. Tom's doesn't list the timings for the memory, but at 4K the CPU shouldn't matter as you can practically use a Celeron and get the same results as the top Intel gaming CPU.

It's not as though a difference is completely unexpected. SotTR was tested in 9 reviews. 4 of them had the 6800 XT at 91 - 93% of the 3080. 2 reviews had it at 97-98%, and the remaining 3 had the 6800 XT at 101 - 102% over the 3080. However those aren't too wide and could be chalked up to which part of the game is being tested.

The swing between Strange Brigade is much larger (and even a little larger than you would expect since it's not as simple as just taking the difference of the two numbers) and once again it's difficult to imagine that any minor differences in CPU (they all test using similarly high-end Intel or AMD CPUs anyways) would create that result since those differences are basically negated at 4K

 

[vissarix]

Another overhyped train derailed? Some wise posters over here telling us how Nvidia rushed the RTX30 series because they had an inferior product , and how the RX6800XT was going to be much faster than the RTX3080, but as always reality shows up

 

[psolord]

^Dude can you quit with the trolling already? How old are you seriously?

Having spent the last 6 years on Nvidia cards, I am certainly not an avid AMD lover, but it seems you forgot other important graphs of that review.

They are selling you a similarly performing card within 5%, at 50$ less and with 60% more VRAM. WTF more you want?

And who said that the 6800XT was going to be faster than the 3080?

 

[vissarix]

^Dude can you quit with the trolling already? How old are you seriously?

Having spent the last 6 years on Nvidia cards, I am certainly not an avid AMD lover, but it seems you forgot other important graphs of that review.

They are selling you a similarly performing card within 5%, at 50$ less and with 60% more VRAM. WTF more you want?

And who said that the 6800XT was going to be faster than the 3080?

For someone not being an ''AMD lover'' you got seriously butthurt and came back with a weak response and personal attacks, So you would rather buy a card that is 3% more efficient or 5% faster? Guess what? No one in earth is going to consider that 3% of more efficiency in this case..

 

[Antey]

Another overhyped train derailed? Some wise posters over here telling us how Nvidia rushed the RTX30 series because they had an inferior product , and how the RX6800XT was going to be much faster than the RTX3080, but as always reality shows up

uhm, vissarix, i think you are confused, wccftech would be that way

-->

Wccftech

We bring you the latest from hardware, mobile technology and gaming industries in news, reviews, guides and more.

wccftech.com

 

[Qwertilot]

There were some fairly silly things dreamed of, I think In both directions! The cards are priced quite well against each other as you'd expect given the supply/demand situation.

The situation with notebooks could be rather interesting. Looks quite balanced overall (the 3080/90 do drop some perf/watt to get huge compute in). Which is a huge advance for AMD but not going to turn the market over.

 

[Glo.]

Another overhyped train derailed? Some wise posters over here telling us how Nvidia rushed the RTX30 series because they had an inferior product , and how the RX6800XT was going to be much faster than the RTX3080, but as always reality shows up
View attachment 34214

Why don't you also include power draw figures, and 1080p and 1440p charts?

Because simply AMD has superior products, and you do not want that to disprove your agenda.

 

[Leadbox]

Why don't you also include power draw figures, and 1080p and 1440p charts?

Because simply AMD has superior products, and you do not want that to disprove your agenda.

Like all the other rabid nvidia fanboys, he's hiding behind 4K and RT results now. Those are the only metrics that matter to them now.

 

[vissarix]

Why don't you also include power draw figures, and 1080p and 1440p charts?

Because simply AMD has superior products, and you do not want that to disprove your agenda.

Enjoy

 

[CastleBravo]

Why don't you also include power draw figures, and 1080p and 1440p charts?

Because simply AMD has superior products, and you do not want that to disprove your agenda.

I wouldn't get too carried away. IMO, AMD and NV have both "released" excellent GPUs this generation. I am still leaning towards a 3080, but if CP2077 ends up working decently well with the 6800 XT, and the 6800 XT is the only one available for purchase, I will go team red.

 

[leoneazzurro]

Enjoy
View attachment 34237View attachment 34238

Maybe you should include also the tests of Computerbase on new titles, that were not included in that average but have a separate table. Should I?

FPS-Unterschiede RX 6000 vs. RTX 3000

Spiel (3.840 × 2.160)	6800 XT vs. 3080	6800 vs. 3070
Assassin's Creed: Valhalla	106 %	114 %
COD Black Ops: Cold War	99 %	113 %
Dirt 5	120 %	124 %
Mafia: Definitive Edition	83 %	99 %
Serious Sam 4	103 %	113 %
Star Wars: Squadrons	106 %	117 %
Watch Dogs: Legion	105 %	115 %
Siege AMD/Nvidia	5/2	6/1
3080 und 3070 normiert auf 100 Prozent

AMD Radeon RX 6800 und RX 6800 XT im Test: Die Taktraten, Benchmarks in Full HD, WQHD sowie Ultra HD und Raytracing

AMD Radeon RX 6800 XT im Test: Die Taktraten, Benchmarks in Full HD, WQHD sowie Ultra HD und Raytracing / Testsystem und Testmethodik

www.computerbase.de

 

[Kenmitch]

Enjoy

I don't understand why charts are posted without at least a link to the review. What are you hiding?