You can sign up to get a daily email of our articles, see the Mailing List page!

Today AMD formally revealed the next-generation Radeon GPUs powered by the RDNA 2 architecture and it looks like they’re going to thoroughly give NVIDIA a run for your money.

What was announced: Radeon RX 6900 XT, Radeon RX 6800 XT, Radeon 6800 with the Radeon RX 6800 XT looking like a very capable GPU that sits right next to NVIDIA's 3080 while seeming to use less power. All three of them will support Ray Tracing as expected with AMD adding a "high performance, fixed-function Ray Accelerator engine to each compute unit". However, we're still waiting on The Khronos Group to formally announce the proper release of the vendor-neutral Ray Tracing extensions for Vulkan which still aren't finished (provisional since March 2020) so for now DirectX RT was all they mentioned.

Part of the big improvement in RDNA 2 comes from what they learned with Zen 3 and their new "Infinity Cache", which is a high-performance, last-level data cache they say "dramatically" reduces latency and power consumption while delivering higher performance than previous designs. You can see some of the benchmarks they showed in the image below:

As always, it's worth waiting on independent benchmarks for the full picture as both AMD and NVIDIA like to cherry-pick what makes them look good of course.

Here's the key highlight specifications:

  RX 6900 XT RX 6800 XT RX 6800
Compute Units 80 72 60
Process TSMC 7nm TSMC 7nm TSMC 7nm
Game clock (MHz) 2,015 2,015 1,815
Boost clock (MHz) 2,250 2,250 2,105
Infinity Cache (MB) 128 128 128
Memory 16GB GDDR6 16GB GDDR6 16GB GDDR6
TDP (Watt) 300 300 250
Price (USD) $999 $649 $579
Available 08/12/2020 18/11/2020 18/11/2020

You shouldn't need to go buying a new case either, as AMD say they had easy upgrades in mind as they built these new GPUs for "standard chassis" with a length of 267mm and 2x8 standard 8-pin power connectors, and designed to operate with existing enthusiast-class 650W-750W power supplies.

There was a big portion of the event dedicated to DirectX which doesn’t mean much for us, but what we’ve been able to learn from the benchmarks shown is that they’re powerful cards and they appear to fight even NVIDIA’s latest high end consumer GPUs like the GeForce 3080. So not only are AMD leaping over Intel with the Ryzen 5000, they’re also now shutting NVIDIA out in the cold too. Incredible to see how far AMD has surged in the last few years. This is what NVIDIA and Intel have needed, some strong competition.

How will their Linux support be? You're probably looking at around the likes of Ubuntu 21.04 next April (or comparable distro updates) to see reasonable out-of-the-box support, thanks to newer Mesa drivers and an updated Linux Kernel but we will know a lot more once they actually release and can be tested.

As for what’s next? AMD confirmed that RDNA3 is well into the design stage, with a release expected before the end of 2022 for GPUs powered by RDNA3.

You can view the full event video in our YouTube embed below:

YouTube Thumbnail
YouTube videos require cookies, you must accept their cookies to view. View cookie preferences.
Accept Cookies & Show   Direct Link

Feel free to comment as you watch as if you have JavaScript enabled it won't refresh the page.

Additionally if you missed it, AMD also recently announced (October 27) that they will be acquiring chip designer Xilinx.

Article taken from GamingOnLinux.com.
38 Likes, Who?
We do often include affiliate links to earn us some pennies. We are currently affiliated with GOG, Humble Store and Paradox Interactive. See more here.
About the author -
author picture
I am the owner of GamingOnLinux. After discovering Linux back in the days of Mandrake in 2003, I constantly came back to check on the progress of Linux until Ubuntu appeared on the scene and it helped me to really love it. You can reach me easily by emailing GamingOnLinux directly.
See more from me
150 comments
Page: «12/15»
  Go to:

mirv 29 Oct, 2020
View PC info
  • Supporter Plus
Quoting: lunix
Quoting: mirv
Quoting: lunixOSS is a tool of Free Software which is political - just as you described. You're literally talking about ideological circumstances and not about actual value.

But those circumstances do have actual value to people. This is the point I think you're not understanding.

It's really a shame that you're so caught up on apparently telling people they are wrong in this, because there are other points that might be nice to discuss. Can we just drop the whole thing and concentrate on what these cards deliver in tangible features, or lack in tangible features?


So you changed your mind again? I'm not telling them they are wrong about FOSS - what I'm telling them IF they want a future with FOSS in it then we need to concentrate on features first to gain momentum. Linux people just constantly complain about the lack of support but it's a catch 22: small community = little to no support. To fix that we need to "bait" outsiders. Once we have a big audience we will get the support and FOSS. Obviously the linux gaming community would be much smaller if linux wouldn't have amd, nvidia, vulkan or proton because people want stuff. People also try to tell me what is "smart" but I'm going to disagree and say GPUs have lots of parameters, pros and cons. I don't need to support a company lol - the companies need to deliver value to us.

My main concern is if there's a real desire to increase users, that FOSS principles (user freedoms) will be ignored - and once ignored or removed, companies will not want to return them. So I personally rather the approach of slowly improve and add features - make it useful, and users will come. It still needs backing of some kind (large community or a company) to improve and add features of course - basically the kind of work Valve is funding is how it should be done, as I see it.

On that note, I would like companies to deliver experimental features into FOSS drivers. Get them into the hands of a wider range of developers, get their feedback. Proprietary extensions to OpenGL and Vulkan, for example, are almost never visible in FOSS drivers and that is a real shame in my opinion.

Quote
Quoting: mirvEven something so simple as: how much power do they draw when idle, what kind of heat do they generate, what noise is produced from the fans running. Power consumption is actually another very important metric of mine, and I don't go out buying the latest & greatest for that reason. I wait a bit to see what comes out that might not use more electricity than my entire current system on full load. Bills to pay and all that.

Power consumption is also important to me because of two factors: I don't really want to waste more electricity than needed and because it decides how much heat it generates which affects its performance.

I'm really curious if the NZXT Kraken G12, Morpheus II or Arctic Accelero IV aftermarket coolers will be compatible with these because they're usually far better than stock coolers. The G12 allowed me to use an AIO and I want to keep using it. The RTX 3000 line is not compatible with the G12(unless you're into crafting) and both the VRM and the GDDR6X are slotted around the chip which makes it hard to cool them. That's architecturally better for everyone but people like me who use aftermarket coolers.

I'm really interested in their Smart Access Memory tech and how is it going to work on linux because I'm buying a Zen 3 next month(sweet single-core performance).

Indeed, I was wondering about that smart access memory thing, and if they would do anything with it on GNU/Linux. Did I read somewhere that it might just be more convenient mapping of GPU memory ranges? I'll have to go back and find more info about that - AMD are in a good position to (and I can't believe I'm using this kind of wording) provide a synergy of GPU and CPU that nobody else can yet do. Intel maybe down the road, we'll have to see what comes of it, but AMD have more experience in this kind of area.
Mohandevir 29 Oct, 2020
What's the bottom line of all this Lunix? What are you trying to convince us?
Feels like this discussion as lasted way too long.

You got your opinion, fine... Stick to it. Your point of view is your own and has as much value as any other from anyone here. It's all biased in some way by our own personnal experiences that differs from one person to another...

Do you want to have an open discussion or you just want to show us all that you are right? At this point, it's beginning to look like trolling.

Please, be on your way, there is nothing to see...
lunix 29 Oct, 2020
Quoting: mirvIndeed, I was wondering about that smart access memory thing, and if they would do anything with it on GNU/Linux. Did I read somewhere that it might just be more convenient mapping of GPU memory ranges? I'll have to go back and find more info about that - AMD are in a good position to (and I can't believe I'm using this kind of wording) provide a synergy of GPU and CPU that nobody else can yet do. Intel maybe down the road, we'll have to see what comes of it, but AMD have more experience in this kind of area.

What I read is that when the CPU needs to access VRAM they use an indirect buffer but SMA will make it possible for it to directly address those pages through the PCIe. It's a bit anti-competitive but I expect this to appear on the other chips too because consoles also started to share the VRAM as the system RAM. I wait for the day when the system and GPU memory will be unified and modularized.
lunix 29 Oct, 2020
Quoting: MohandevirWhat's the bottom line of all this Lunix? What are you trying to convince us?
Feels like this discussion as lasted way too long.

You got your opinion, fine... Stick to it. Your point of view is your own and has as much value as any other from anyone here. It's all biased in some way by our own personnal experiences that differs from one person to another...

Do you want to have an open discussion or you just want to show us all that you are right? At this point, it's beginning to look like trolling.

Please, be on your way, there is nothing to see...

You're overthinking it, I just wanted to warn people to carefully evaluate features because if they don't do that they can get burned and then a few months later they'll start to rant how they'll never buy anything from $gpu-company again. And features aren't going to develop themselves just because you buy another product. Nothing special, but it became political. All concerns are valid, the point is to be aware of them.
mirv 29 Oct, 2020
View PC info
  • Supporter Plus
Quoting: lunix
Quoting: mirvIndeed, I was wondering about that smart access memory thing, and if they would do anything with it on GNU/Linux. Did I read somewhere that it might just be more convenient mapping of GPU memory ranges? I'll have to go back and find more info about that - AMD are in a good position to (and I can't believe I'm using this kind of wording) provide a synergy of GPU and CPU that nobody else can yet do. Intel maybe down the road, we'll have to see what comes of it, but AMD have more experience in this kind of area.

What I read is that when the CPU needs to access VRAM they use an indirect buffer but SMA will make it possible for it to directly address those pages through the PCIe. It's a bit anti-competitive but I expect this to appear on the other chips too because consoles also started to share the VRAM as the system RAM. I wait for the day when the system and GPU memory will be unified and modularized.

Ah yes, that was it. Yeah AMD have been all about a better memory architecture between CPU and GPU, even since the first APUs started to appear - and I can see desktops and consoles kind of converging (architecturally speaking) in the future. It's already started I suppose.

Real catch is going to be unified memory from a user perspective, but keeping it fast for both CPU and GPU requirements. SMA might be a step in that direction.
Hmm...Vulkan already does have the infrastructure in place to treat memory in this fashion too, so that's useful.
Cyril 29 Oct, 2020
Quoting: lunixThat is pretty much the question of support. You can try older drivers for older nvidia cards, but using older cards is pointless because they will just waste electricity and barely deliver any performance - that is why companies stop supporting them.

What... Producing new cards is a more waste of electricity than using older cards. In this domain, replacing the components is worse than using older ones. But I assume you care only about your electricity consumption, right?
But you assume that everybody just want to play the big recent games while maintaining great performance, in this case sure it's pointless to use old cards.
Using older cards is pointless for you, because you only seem to care about performance (and features, we know).
Gaming doesn't mean for everybody "AAA gaming", and it's particularly true for Linux users, as other said: we could just stick with Windows.
If someone is happy using an old card because it's just works fine, and one day Nvidia (or other) decide to not support it anymore, it's still an issue. This is a waste, throwing components that still works because people have to buy new ones is nonsense for me...

I read all the thread, and, lunix, what a guy you are.
At GOL we're used to see guys like you, from time to time.
Mohandevir 29 Oct, 2020
Quoting: lunix
Quoting: MohandevirWhat's the bottom line of all this Lunix? What are you trying to convince us?
Feels like this discussion as lasted way too long.

You got your opinion, fine... Stick to it. Your point of view is your own and has as much value as any other from anyone here. It's all biased in some way by our own personnal experiences that differs from one person to another...

Do you want to have an open discussion or you just want to show us all that you are right? At this point, it's beginning to look like trolling.

Please, be on your way, there is nothing to see...

You're overthinking it, I just wanted to warn people to carefully evaluate features because if they don't do that they can get burned and then a few months later they'll start to rant how they'll never buy anything from $gpu-company again. And features aren't going to develop themselves just because you buy another product. Nothing special, but it became political. All concerns are valid, the point is to be aware of them.

Agreed. And when you say "$gpu-company" we can add "from chipset to manufacturer". Not all AMD or Nvidia GPUs are born equal (MSI, ASUS, Gigabyte, Sapphire, Power Color, etc...) Making the difference between a manufacturer's bad "QA decision" and a chipset flaw is also important. Been there a couple of times.


Last edited by Mohandevir on 29 October 2020 at 4:00 pm UTC
lunix 29 Oct, 2020
Quoting: Cyril
Quoting: lunixThat is pretty much the question of support. You can try older drivers for older nvidia cards, but using older cards is pointless because they will just waste electricity and barely deliver any performance - that is why companies stop supporting them.

What... Producing new cards is a more waste of electricity than using older cards. In this domain, replacing the components is worse than using older ones. But I assume you care only about your electricity consumption, right?
But you assume that everybody just want to play the big recent games while maintaining great performance, in this case sure it's pointless to use old cards.
Using older cards is pointless for you, because you only seem to care about performance (and features, we know).
Gaming doesn't mean for everybody "AAA gaming", and it's particularly true for Linux users, as other said: we could just stick with Windows.
If someone is happy using an old card because it's just works fine, and one day Nvidia (or other) decide to not support it anymore, it's still an issue. This is a waste, throwing components that still works because people have to buy new ones is nonsense for me...

I read all the thread, and, lunix, what a guy you are.
At GOL we're used to see guys like you, from time to time.

New cards are always significantly more efficient but we are talking about 10+ year old cards which are poorly supported and provide a very bad experience. Do you care about efficiency at the low-end? Buy APUs. Yeah, most people will buy new cards when their older ones can no longer perform properly. The amount of electricity you will waste with an inefficient architecture will be many times larger than the electricity used to create a new GPU. Old cards can be recycled and the manufacturing process is nothing compared to the daily waste. Don't want "AAA" gaming? Don't buy premium cards then.

The reason why companies drop support is because maintaining drivers will become a pointless chore with an older architecture and just because a handful of users are still on it, it won't justify the costs - including the electricity what the company would use to test the GPU with the latest drivers.


Last edited by lunix on 29 October 2020 at 4:11 pm UTC
libgradev 29 Oct, 2020
Quoting: Liam DaweI do have to admit personally, even as a typical NVIDIA fan that AMD are looking mighty tempting.

Join us

6900 XT or the 6800 XT - hmmm....
jens 29 Oct, 2020
View PC info
  • Supporter
Quoting: libgradev
Quoting: Liam DaweI do have to admit personally, even as a typical NVIDIA fan that AMD are looking mighty tempting.

Join us

6900 XT or the 6800 XT - hmmm....

Yeah same here. I'm actually quite happy with my GTX 1080, but due to VR I want to go for something more powerful soon. I don't mind that the Nvidia driver is closed source, I have to use dkms also for an out of tree gaming wheel driver. That said not having async reprojection* or being able to switch to wayland might be the reason to go for one of those cards.

Lets see how the actual launch and support looks like...

* I don't know if it is a hack or not, but it really does serves its purpose.


Last edited by jens on 29 October 2020 at 6:44 pm UTC
While you're here, please consider supporting GamingOnLinux on:

Patreon, Liberapay or PayPal Donation.

We have no adverts, no paywalls, no timed exclusive articles. Just good, fresh content. Without your continued support, we simply could not continue!

You can find even more ways to support us on this dedicated page any time. If you already are, thank you!
Login / Register

Or login with...
Sign in with Steam Sign in with Twitter Sign in with Google
Social logins require cookies to stay logged in.

Livestreams & Videos
Community Livestreams
Latest Forum Posts