Is it possible that the RX 6800XT is faster than the 3090? - Page 3 - Video Cards - HWzone Forums
Skip to content
  • Create an account
  • About Us

    Hello Guest!

     
    Please note - in order to participate in our community, comment and open new discussions, you must join as a registered member.

    Our members enjoy many advantages, including the ability to participate in discussions, enjoy raffles and promotions for members of the site, and receive our weekly content directly by email.

    Do not like being harassed by email? You can register for the site but do not submit your registration to the weekly email updates.

Is it possible that the RX 6800XT is faster than the 3090?


Recommended Posts

Wait wait as they say - Shuya Shuya : beerchug: 

We are now doing a suite of all the measurements Lior made in the 0 day review of the launch, as a reference to the reference we

Want to measure -> before And after

 

Then we will all see and learn together, how much percentage (and nominal) performance improvement these cards achieve when you rush them

To the potential they are capable of. And they are hugely capable. Reminiscent of the good old days of the HD7950 from 2011,

Which marketed at an anemic frequency of only 800mhz but made 1150MHz up to 1200 in practice.

 

As it seems at the moment and with the full precaution required, we are in the middle of the measurements yes, but see about 15-20% typical improvement

In games (except for exceptions seen below). Therefore there is an optimistic basis to assume at this time, that in most measurements a gap in favor will narrow

The 6800xt is in a hurry. Patience friends will lock everything up.

 

** And titles you ask for beyond that, just say what and we will try to get. As a service to the community and friends with wholehearted love.

Edited By nec_000
Post Reply Direct Link To This Post
Share on other sites
  • תגובות 297
  • Opens on
  • Last comment

Prominent participants in the discussion

Prominent participants in the discussion

Popular responses

What closes with these messages? What do they contribute? Porn of box card boxes? Not so clear to me ...

If so, at a good time we finished from the first end where we measure the eight titles we originally tested on 6800xt so we have a reference for reference and comparison against each other. We are very happy that we insisted on finding a good 3080 card, and were not satisfied

Do not understand this narrative of "Fine Wine". I do not know in which communities you hang out, in the ones I hang out in it is mostly such a network joke. You could call it "drivers that get better with time" ... you could also call it bad drivers

Attached Images

Good,

Now we come to the red dead redemption test, as I mentioned we are going through a title title made by Lior in the source review, and we have now finished measuring

Also the title is. lanzar achieves 116FPS compared to 97.6 by default.

 

Settings of course identical to everything on ultra, note for the avoidance of doubt:

3.thumb.jpg.61d0b14a466af9c682ed3f45da17e3fe.jpg

 

The measurement, 115.928 - that it will consume 116FPS:

4.thumb.jpg.71fc9c6f3d083ec2f09d15f4382886a3.jpg

 

And this is the reference to the reference without . This is an improvement of 20% and of course a significant gap opening above 3090:

RDR2-2.thumb.png.62fb1a075f3c5cfc3902bd6658d50ec4.png

 

 

In consultation with lanzar we decided that at the end a summary graph would be made, showing the performance of the 6800XT by default, as opposed to the performance

After OC as obtained in measurements. The percentage improvement is calculated as a percentage. As a reference we will of course place the 3080 and the 3090

For a comparative impression. Will be fascinating - promising.

Edited By nec_000
Post Reply Direct Link To This Post
Share on other sites

 

We got to the Metro exam,

 

The result obtained in the same one-to-one settings as Lior ran is 118.4FPS

8.thumb.jpg.78e3b9b146ee9a2574c4f822d207a929.jpg

 

And this is the reference for reference - as mentioned we have to go from 105 to 118.4 FPS

Metro-1.thumb.png.63c44b7b4fafa0bb92a894009c861fd1.png

 

 

Edited By nec_000
Post Reply Direct Link To This Post
Share on other sites

Now the exact same test failed Lior on 3dmark timespy in the same settings one by one.

One of the things we noticed is that in the synthetic Benz in question, the load on the card is more unusual than in regular games,

What is quite predictable, however, is that this feature brings the power consumed to the 320w ceiling in which the card is enclosed, intentionally

על ידי Of course, then the card starts to drop in frequency and no longer maintains 2.7Ghz plus as happens in games, and converges

More to 2.5Ghz. In this case where the full theoretical conversion given (in games) is not achieved, and will require a BIOS mode

In order to open this ceiling in the card (which removes warranty of course). This is not something that will be done at this time, and will be considered

Performed later (if reported and measured again).

 

Given a 320W ceiling the result obtained is 19287 points at the graphics level:

3dtimespy.thumb.jpg.93454c7e084d19e59b5f615342eac89e.jpg

 

Compared to the reference to the relational comparison which shows 17521 points.

This is an improvement of only about 10% (because the frequency did not increase as required up to 2.7ghz and reached the current ceiling in the region of 2.5ghz)

3dmark.png.14c73aaa26bf0ad166de2575712c85d7.png

Edited By nec_000
Post Reply Direct Link To This Post
Share on other sites

Say you'm serious? You compare Tested on two different systems and under different conditions.
The images you upload here have no meaning unless you put in the exact same system and under the same conditions 3090 and run the tests.
Luckily there is no need for it because there are dozens of reviewers who have invested a lot more time and resources in it and you are more than welcome to go read what they are writing.

And for the sake of rasterization in 1080P the 6800 XT is indeed better, but the second you run DLSS or RT it is not in the same neighborhood even and it loses at higher resolutions.
(In fact if DLSS and RT operators get both more performance and a better picture with Noydia cards)

 

Edited By Moon-Mage
Post Reply Direct Link To This Post
Share on other sites

What was said in the discussion from what I understood is that the speed of memories should not be increased as it actually hurts performance.

Accordingly I would be very happy to read the rest of the thread.

It seems that this is information that needs to be released beyond the boundaries of the forum 

6800XT water cooled About 2743 gives improvement of single FPS barely in techpowerup in RDR2. 

But there too the memory is rushed. In fact almost everywhere they did an OC test on the card, they also speed up the memory.

 

There is a number difference here that greatly exceeds any range of error and standard deviation. 

 

And yes, in the end the information needs to be re-processed by someone who has already officially performed the test.

The green cards may also need to be checked under OC without tampering with memory.

 

Yes, apparently those who want RT have not even considered this card at all.

And yes, DLSS is essential. 

But the information presented is no less significant, especially in terms of the significance of mental fixation that may have stuck to hardware testing sites. 

 

Post Reply Direct Link To This Post
Share on other sites

The test systems hardly change the average performance when they are of a similar caliber, such as the 10900K or 5900X, certainly not at high resolutions at full settings, where what plays the most role is the GPU.

 

I do not know what frightened / stressed you. We bring you results so that if you decide to buy an AMD card, you will know what it is capable of soon, as it seems AMD released cards with a significantly lower working frequency than the cores (most of them at least) are capable of working, the opposite of For example, that some of the cards she released collapsed even at a voltage / frequency, which forced them to release With a slightly lower running frequency - that is, they are pretty much on the edge in this respect.

RAY TRACING is nice, but finishes the performance (if we put the DLSS aside) even if we use a monster like the 3090. We are not there yet, what saves the performance in working with RAY TRACING is the DLSS, which you get a response to DLSS from In the upcoming / next dirver, and according to them he will support all the games and not just the games that have been planned for this in advance.

Edited By Lanzar
Post Reply Direct Link To This Post
Share on other sites
Quote of yoavke

What was said in the discussion from what I understood is that the speed of memories should not be increased as it actually hurts performance.

Accordingly I would be very happy to read the rest of the thread.

It seems that this is information that needs to be released beyond the boundaries of the forum 

6800XT water cooled Asus About 2743 gives improvement of single FPS barely in techpowerup in RDR2. 

But there too the memory is rushed. In fact almost everywhere they did an OC test on the card, they also speed up the memory.

 

There is a number difference here that greatly exceeds any range of error and standard deviation. 

 

And yes, in the end the information needs to be re-processed by someone who has already officially performed the test.

The green cards may also need to be checked under OC without tampering with memory.

 

Yes, apparently those who want RT have not even considered this card at all.

And yes, DLSS is essential. 

But the information presented is no less significant, especially in terms of the significance of mental fixation that may have stuck to hardware testing sites. 

 

 

Exactly Yoav. From there everything was born. Our friend Lanzar got the card so the business was born.

He saw significantly better performance than he sees in the reviews and rubbed his eyes. Did not understand what was happening.

Includes reviews / demonstrations when speeding up the card to the 2.7ghz area. So he looked for what was different and came up with two recurring things

about themselves:

 

The first is that in the speeds they did online for 2.7, they also inadvertently speeded up VRAM - leading to a typical drop to 10%

In FPS. Those in a hurry did not think for a moment that one should check with and without, and understand what causes it. Then he will recreate their results

With the speed of VRAM, and so came up with the issue basically. That they are making a mistake.

 

The second thing he found, that most still work with platforms of And demonstrate with them, and he has one under his arm

10900K, and note that in some cases (not all) but in some, you see better performance when working on the Risen platform

5000. What makes sense - this is a more powerful processor and a platformer with more BUS.

 

At this point he revealed to me the findings (he and I are close friends) so I told him not to keep such information just for himself,

But need to share with friends and Lior. And so we did.

 

now,

For testing he has 3070 currently at home so he tried last night to see if the performance improvement obtained in operation

6800XT on Risen 5000 instead of on Recurring (which is our premise) also for the 3070.

As it turns out, our assumption came out wrong. The 3070 was not found to receive a similar performance boost over the Risen 5000 in those locations

That 6800XT did get. 

 

hearing,

That the improvement in this matter, of going beyond the Risen 5000 platform, with the full necessary care and as it appears on the face of it in the test

The quick thing he did, is that it only affects the case of RDNA2 architecture. They probably optimized something there and there may be

A connection to the new stint of the CPU's direct access to VRAM on the card, the one they boasted about in the novelty in question.

This may be one of the most contributing factors in the move to the Risen 5000 and this is currently our current assumption.

 

We will continue to check the other titles and measurements made by Lior and in the end we will put everything on a neat graph.

 

** We would of course like to make parallel measurements with card 3090 (if we get, very difficult to locate) to produce our own reference,

Although Lior did an excellent job and can certainly be used with his work as a reference for comparison.

Or we will use the caveman who has 3090 and also Risen 5000 which is the ideal of course, and we will see what he can contribute of his time and measure

For us as much as he can help us. It's a lot of work and downloads and what not ... holding fingers that can help 🙏

Edited By nec_000
Post Reply Direct Link To This Post
Share on other sites

Join the discussion

You can then join the discussion and then join our community. If you already have an account with us, please Log in now To comment under your username.
Note that: The comment will appear to the surfers after approval by the board management team.

guest
Add a comment

×   The content pasted is with formatting.   Remove formatting

  Only 75 emoji are allowed.

×   Your link has been automatically assimilated.   Show as regular link

×   Your previous content has been automatically restored.   Clear all

×   You can not paste images directly. Upload or insert images from URL.


×
  • Create new ...

At the top of the news:

In June everything will work out? The MWC 2021 exhibition begins with a Corona-adapted version

Instead of a grandiose mobile celebration in Barcelona - we get a relatively small event in Shanghai this week, and plans for a European comeback this coming June The MWC (Mobile World Congress) was the first notable technological event to suffer last year's coronation, with naive restrictions in the first stage, Complete cancellation of the physical event afterwards - and at the same time promises that next year everything will return to normal, when it is likely that at that point in time no one guessed or predicted that even in 2021 ...

new on the site