AMD Polaris architecture – GCN 4.0

drunkenmaster · 25 Feb 2016 at 01:49

james.miller said:
I am genuinely suprised that you think witcher 3 caused so many problems for anybody, let alone nvidia users. hairworks stuff aside, i thought the game was well received and pretty bug free from a technical standpoint. Certainly there were issues but the vast majority of the patches from CDPR addressed gameplay bugs, not technical issues.

Anyway, the 290x spanking the 780ti is not something i remember, not at 1080p or 4k (which was unplayable on any single card with ultra settings). Certainly looking back at reviews such as Techspots, the 780ti was either just in front or just behind (at unplayable framerates).

So, care clarify either of those statements? You know, with some proof or something.

http://www.guru3d.com/articles-pages/the-witcher-3-graphics-performance-review,6.html

Every single other review I find has the 290x ahead, fairly significantly, for a card that cost significantly less. Kepler performance was dire in Witcher 3 on launch to the tune that Nvidia themselves finally admitted poor Kepler performance and specifically released a Kepler fix driver... which a few fanboys on here think doesn't confirm a kepler performance problem.

There were for kepler problems dozens of threads with thousands of posts on Nvidia forums for around 5-6 months complaining about Kepler performance.

In regards to Nvidia and the game, the game itself had no bugs. Nvidia released like 5 beta's to keep fixing their own problems. Thousands of Nvidia users, threads on every major gaming forum talked about stability problems with all the drivers released for Witcher 3. As with all drivers, not everyone is effected and every new driver will screw up someones system. But the sheer numbers of Nvidia users who on any forum I visit were complaining about problems. It still wasn't close to anyone but it was significantly more than one, acceptable, significant enough for Nvidia to release multiple drivers in a very short space of time trying to fix it.

I and other AMD users had dodgy xfire support(sli wasn't great either afaik) but rock solid single card use on a driver released 4 months prior. The driver for the game specifically was rock solid on stability for the game, the drivers later were rock solid for the game.

flopper · 25 Feb 2016 at 09:34

drunkenmaster said:
http://www.guru3d.com/articles-pages/the-witcher-3-graphics-performance-review,6.html

Every single other review I find has the 290x ahead, fairly significantly, for a card that cost significantly less. Kepler performance was dire in Witcher 3 on launch to the tune that Nvidia themselves finally admitted poor Kepler performance and specifically released a Kepler fix driver... which a few fanboys on here think doesn't confirm a kepler performance problem.

There were for kepler problems dozens of threads with thousands of posts on Nvidia forums for around 5-6 months complaining about Kepler performance.

In regards to Nvidia and the game, the game itself had no bugs. Nvidia released like 5 beta's to keep fixing their own problems. Thousands of Nvidia users, threads on every major gaming forum talked about stability problems with all the drivers released for Witcher 3. As with all drivers, not everyone is effected and every new driver will screw up someones system. But the sheer numbers of Nvidia users who on any forum I visit were complaining about problems. It still wasn't close to anyone but it was significantly more than one, acceptable, significant enough for Nvidia to release multiple drivers in a very short space of time trying to fix it.

I and other AMD users had dodgy xfire support(sli wasn't great either afaik) but rock solid single card use on a driver released 4 months prior. The driver for the game specifically was rock solid on stability for the game, the drivers later were rock solid for the game.

Of course, AMD does things right.

This generation its called Polaris graphics card done right for a brighter gaming future.

The migration from 980ti owners have started to order and buy Polaris.

Mtom · 25 Feb 2016 at 10:45

flopper said:
Of course, AMD does things right.

This generation its called Polaris graphics card done right for a brighter gaming future.

The migration from 980ti owners have started to order and buy Polaris.

Problem is with the difference in marketing. NV has great marketing, and AMD has meh.

james.miller · 25 Feb 2016 at 11:05

drunkenmaster said:
http://www.guru3d.com/articles-pages/the-witcher-3-graphics-performance-review,6.html

Every single other review I find has the 290x ahead, fairly significantly, for a card that cost significantly less. Kepler performance was dire in Witcher 3 on launch to the tune that Nvidia themselves finally admitted poor Kepler performance and specifically released a Kepler fix driver... which a few fanboys on here think doesn't confirm a kepler performance problem.

Guru3d had the 290x leading by 6fps at 1080p and 3fps at 4k, average fps, which is basically 14%. So ill give you that, it was faster, but noticable? both are unplayable at 4k and 1080p...maybe just.

Both slower than the hobbled Maxwell of course, showing there was of course an issue with kepler which nobody should be denying. However, you are being pretty careful with your words and inferring that there were problems across the board for nvidia users when actually no there was not, it was kepler users specifically. It was kepler issues that had nvidia releasing multiple drivers, none of which affected Maxwell in any meaningful way.

So no, this isnt a case of nVidia fell over and AMD lead the way with the light shining out of their backside - it's a case of nvidia dropping the ball with Kepler and not just in witcher 3 but pretty much across the board. This was not unique to witcher 3 and it didnt affect nvidia best selling card

I and other AMD users had dodgy xfire support(sli wasn't great either afaik) but rock solid single card use on a driver released 4 months prior. The driver for the game specifically was rock solid on stability for the game, the drivers later were rock solid for the game.

As a 970 user so did I

drunkenmaster · 25 Feb 2016 at 11:56

No, stability was across the board, Kepler had additional performance issues. The multiple beta's were to address stability full stop. It took months, and months, and months of Nvidia users complaining on Nvidia forums to get a Kepler fix driver which was an entirely different driver addressing that one problem. The half dozen drivers repeatedly attempting to fix widespread problems had nothing to do with the Kepler performance issue.

They released a single Kepler performance fix... only after multiple websites actually wrote articles and more than the usual forums were complaining about Kepler performance and magically after 6 months of problems they had a fix in a couple of days.

Overlag · 25 Feb 2016 at 17:58

http://www.anandtech.com/show/10067/ashes-of-the-singularity-revisited-beta

If only DX12 support would happen faster. Most of my games still run on DX9!

D.P. · 25 Feb 2016 at 18:08

humbug said:
lol @

i have to say tho i do find it very very odd that the 980 is apparently 30% faster than the 970 with only 23% more CUDA Cores, thats impossible, its way over 100% scaling. one CUDA core does not make more than one CUDA core.
And we know at best core scaling is 0.7 in 1.

There is more than the number of CUDA cores to account t for thoufh, so as reduced L2 cache

james.miller · 25 Feb 2016 at 18:30

and ROPs

humbug · 25 Feb 2016 at 18:41

It doesn't normally make that much of a difference, 10 to 15% is the norm, that is 30%, 2 to 3 times as much.

LoadsaMoney · 26 Feb 2016 at 15:12

CAT-THE-FIFTH · 27 Feb 2016 at 18:21

This is quite interesting:

https://translate.google.co.uk/tran...Radeon-Geforce-Future-GPU-1187323/&edit-text=

The demo is just over 4000 bytes,and seems to be mostly compute based - AMD cards do very well in it!!

humbug · 27 Feb 2016 at 18:37

CAT-THE-FIFTH said:
This is quite interesting:

https://translate.google.co.uk/tran...Radeon-Geforce-Future-GPU-1187323/&edit-text=

The demo is just over 4000 bytes,and seems to be mostly compute based - AMD cards do very well in it!!

You can get it here

720P appears to be locked to 30 FPS, but (For future GPU's 1080P) i got 8 to 20 FPS.

Who wants to make the benchmark thread?

Dygaza · 27 Feb 2016 at 18:38

Benchmark thread for 4kb demo?

humbug · 27 Feb 2016 at 18:39

Dygaza said:
Benchmark thread for 4kb demo?

Hey don't be critical

, try it out.

Jono8 · 27 Feb 2016 at 18:47

humbug said:
It doesn't normally make that much of a difference, 10 to 15% is the norm, that is 30%, 2 to 3 times as much.

I don't know, I think it depends on the game/benchmark. I remember when I bought my 980 and scouring various reviews that generally the 980 was 15-20% quicker on average.

Here is another 30% plus one though, Shadow of Mordor above 1080p

http://www.anandtech.com/bench/product/1595?vs=1442

Generally you are talking between 15-20% though (most of the anandtech bench ones are around 16/17%) with a few results falling near 10% and a few falling near 30%

humbug · 27 Feb 2016 at 18:48

Shadow is coincidently? another one of those VRam hogs. ^^^^^
Star Citizen slows right down if i go over 3.5GB. another coincidence?

------

Anyway...

1080P
GTX 970 @ 1500/1950

FRAPS:

2016-02-27 18:43:25 - 2nd_stage_boss
Frames: 2053 - Time: 141250ms -

Avg: 14.535 - Min: 8 - Max: 21

Overlag · 27 Feb 2016 at 19:20

got no way of recording avg fps but i saw a low of 9 and a high of 12 on my 7970 (1000/1375) at 1080p. (msi afterburner log)

MadMatty · 28 Feb 2016 at 10:27

humbug said:
Anyway...

1080P
GTX 970 @ 1500/1950

FRAPS:

2016-02-27 18:43:25 - 2nd_stage_boss
Frames: 2053 - Time: 141250ms -

Avg: 14.535 - Min: 8 - Max: 21

A glimpse of the future!

2nd_stage_boss.1920x1080.20160221.exe

R9 290 @1100/1350

2016-02-28 10:20:58 - 2nd_stage_boss
Frames: 3476 - Time: 149797ms - Avg: 23.205 - Min: 12 - Max: 33

humbug · 28 Feb 2016 at 14:04

MadMatty said:
A glimpse of the future!

2nd_stage_boss.1920x1080.20160221.exe

R9 290 @1100/1350

2016-02-28 10:20:58 - 2nd_stage_boss
Frames: 3476 - Time: 149797ms - Avg: 23.205 - Min: 12 - Max: 33

Yeah. as i predicted.

LoadsaMoney · 1 Mar 2016 at 15:23

AMD could unveil more details about Polaris and R9 Fury X2 this Thursday

http://videocardz.com/58359/amd-could-unveil-more-details-about-polaris-and-r9-fury-x2-this-thursday