FineWine===>FineVinegar FuryX Aging Horrible in 2017

[Kanan]

People change opinions based on the progress of software support. Did I had any idea AMD was not planning to support Fury/X lineup well last year? Hell fucking no. I was still drinking the fucking FineWine cool aid. I was still confident AMD will provide support for SteamVR Async Reprojection. I was still confident AMD will active optimize driver for Fury/X.

What happened in 2017 then? None of that. I bought some new games to play and they run like shit on this GPU.

In case you have been living in the past. 2017 has not been good for the FuryX . Hence why this thread exists. If AMD had better support i would not have made this.

You don't like me disliking AMD GPU. I get it. Nobody is gonna persuade anyone on internet and we all know that. I am done waiting for AMD's never coming feature support and performance optimization. Also i am done debating with you. It will get nowhere.

BTW,go buy yourself a Fury or FuryX and try some 2017 titles for yourself. See if i, along with a whole bunch of hardware reviews are lying.

Just that half of what you're saying is untrue. Only true thing would be that you're unhappy, Fury X is showing normal scaling in 99% of games, the games where it doesn't it doesn't because the architecture simply ISN'T up to. Again. I already wrote that in my first post you ignored above.

I already said, some people just don't accept facts but excuses. I rely on numbers, you and the other guys rely on skewing the interpretation of numbers to whatever is convenient.

Your numbers were already debunked as bullshit. There are multiple better sites on the net with "numbers" that contradict yours. Fury X is doing fine. End of story. This thread is a whiner thread / has no sense whatsoever.

 

[MuhammedAbdo]

Your numbers were already debunked as bullshit. There are multiple better sites on the net with "numbers" that contradict yours. Fury X is doing fine. End of story. This thread is a whiner thread / has no sense whatsoever.

LOL, open the Reddit link in the OP, and show me those multiple sites that contradicts dozens of hard proven links and Youtube comparisons! Your words against this kind of evidence equals nil I am afraid.

 

[Kanan]

LOL, open the Reddit link in the OP, and show me those multiple sites that contradicts dozens of hard proven links and Youtube comparisons! Your words against this kind of evidence equals nil I am afraid.

Strange you're here on TPU and are stupidly ignoring information laid out in front of you:

I'm afraid you're wrong and I'm completely maintaining what I said earlier. Even 1070 has a hard time staying in front of Fury X. Would be untrue if Fury X is bad now, wouldn't it.

Source: https://www.techpowerup.com/reviews/EVGA/GTX_1070_Ti_FTW2/28.html

I also checked every single game. In 2 games it were slow-ish, in 1 game (Watchdogs 2) slow. Performance was in general normal or even great. In many games it was toe-to-toe to the 1070 (which equals a overclocked 980 Ti) and in some even faster. It was almost always faster than the 980 Ti.

Funny then, the OP doubts "FineWine" to be true. It is true, as a matter of fact. Going back to the release of Fury X, Fury X was always *tied* with 980 Ti reference. Now it's faster. Surprise surprise. Reinforces what I said earlier, this thread is bs.

 

[cucker tarlson]

Strange you're here on TPU and are stupidly ignoring information laid out in front of you:

I'm afraid you're wrong and I'm completely maintaining what I said earlier. Even 1070 has a hard time staying in front of Fury X. Would be untrue if Fury X is bad now, wouldn't it.

Source: https://www.techpowerup.com/reviews/EVGA/GTX_1070_Ti_FTW2/28.html

I also checked every single game. In 2 games it were slow-ish, in 1 game (Watchdogs 2) slow. Performance was in general normal or even great. In many games it was toe-to-toe to the 1070 (which equals a overclocked 980 Ti) and in some even faster. It was almost always faster than the 980 Ti.

Funny then, the OP doubts "FineWine" to be true. It is true, as a matter of fact. Going back to the release of Fury X, Fury X was always *tied* with 980 Ti reference. Now it's faster. Surprise surprise. Reinforces what I said earlier, this thread is bs.

The whole point we're making is Fury X having a hard time against the competition in a lot of newly released games, mostly 2H 17, some 1H 17, not TPU charts which are mostly 15/16 titles. If that's not clear after those posts then the situation is really helpless.....

https://www.techpowerup.com/forums/...x-aging-horrible-in-2017.238741/#post-3753872

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-3#post-3754042

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-2#post-3754000

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-6#post-3757791

Same was mentioned in the HWunboxed video testing 2015 cards in 2017 games.

It's hard enough to carry this point across if people "debunk" it by using tpu charts, which are mostly 2015-2016 games. I can't fine ONE 2H 17 game in those TPU charts, and we've had many. Just look at the link to my 1st and last post (purepc and pcgh). "it's right in front of you?" - what is right in front of me ? This is not what we're talking about.

This thread is a shame to human reading comprehension skills. I can't say I'm 100% right on this one but performance charts seem to indicate there is a problem, sometimes quite a big one, people sure as hell can't seem to produce a good counter argument. They can provide an excuse for a high-end card to drop 2 tiers in performance in just one generation, that's for sure. Testing witcher 3 with old and new drivers ? How is that even remotely pertinent ?

 

[Rehmanpa]

Meanwhile I know a guy running pubg at 720p and streaming on a gtx 750 and an i3. If you hate your fury x you could always gift or to a poor pc gamer

 

[ikeke]

The whole point we're making is Fury X having a hard time against the competition in a lot of newly released games, mostly 2H 17, some 1H 17, not TPU charts which are mostly 15/16 titles. If that's not clear after those posts then the situation is really helpless.....

https://www.techpowerup.com/forums/...x-aging-horrible-in-2017.238741/#post-3753872

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-3#post-3754042

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-2#post-3754000

https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-6#post-3757791

Same was mentioned in the HWunboxed video testing 2015 cards in 2017 games.

It's hard enough to carry this point across if people "debunk" it by using tpu charts, which are mostly 2015-2016 games. I can't fine ONE 2H 17 game in those TPU charts, and we've had many. Just look at the link to my 1st and last post (purepc and pcgh). "it's right in front of you?" - what is right in front of me ? This is not what we're talking about.

This thread is a shame to human reading comprehension skills. I can't say I'm 100% right on this one but performance charts seem to indicate there is a problem, sometimes quite a big one, people sure as hell can't seem to produce a good counter argument. They can provide an excuse for a high-end card to drop 2 tiers in performance in just one generation, that's for sure. Testing witcher 3 with old and new drivers ? How is that even remotely pertinent ?

And how exactly is new game on fresh drivers for nvidia but game launch drivers for AMD again a valid and final conclusion on cards performance? (or do you really believe that 980Ti is equal to Vega64 LC?)

Context...again. Versus just throwing numbers that paint the picture you like.

Testing in dynamical software environment is possible if you exclude some variables - so taking games that aren't in beta anymore (as most fresh launch titles have been for the last 10y, IMHO) and comparing them on matured drivers that are equally optimized for different uarch's at said games shows the potential of cards, not some optimization that it's [the card] lacking due to driver/game issues.

Based on what you're and other "finewinegar" crowd has been saying we should have buried Ryzen on first week since it's performance on intel optimized scenarios was worse than comparable Intel chip

 

[MuhammedAbdo]

Testing in dynamical software environment is possible if you exclude some variables - so taking games that aren't in beta anymore (as most fresh launch titles have been for the last 10y, IMHO) and comparing them on matured drivers that are equally optimized for different uarch's at said games shows the potential of cards, not some optimization that it's [the card] lacking due to driver/game issues.

All of the late 2017 games that Fury chocks on are final AAA games. So what is your point exactly?

I'm afraid you're wrong and I'm completely maintaining what I said earlier. Even 1070 has a hard time staying in front of Fury X. Would be untrue if Fury X is bad now, wouldn't it.

Source: https://www.techpowerup.com/reviews/EVGA/GTX_1070_Ti_FTW2/28.html

What a pathetic evidence you have! you mention one site with none of the 2017 games, while the OP gave at least 40 links to at least 20 games, and you have the audacity to consider what you say as evidence. You are splitting straws here.

 

[ikeke]

All of the late 2017 games that Fury chocks on are final AAA games. So what is your point exactly?

What a pathetic evidence you have! you mention one site with none of the 2017 games, while the OP gave at least 40 links to at least 20 games, and you have the audacity to consider what you say as evidence. You are splitting straws here.

FuryX vs 980Ti
http://www.guru3d.com/articles_pages/destiny_2_pc_graphics_performance_benchmark_review,5.html
@1440p it's AVG 58 vs 56
http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,5.html
@1440p it's AVG 44 vs 50
http://www.guru3d.com/articles_page..._pc_graphics_analysis_benchmark_review,5.html
@1440p it's (against GTX1070, 980ti errored out) AVG 73 vs 79
http://www.guru3d.com/articles_page..._pc_graphics_analysis_benchmark_review,5.html
@1440p memory limited - AVG 40 vs 48
http://www.guru3d.com/articles_pages/f1_2017_pc_graphics_performance_benchmark_review,6.html
@1440p it's AVG 71 vs 77
http://www.guru3d.com/articles_pages/prey_pc_graphics_performance_benchmark_review,7.html
@1440p it's AVG 91 vs 93

Yes.

This is choking, definetly.

(again, once playingfield is level and variables are eliminated theres not much between them, but this is difficult to understand if position on performance is taken before all variables are considered)

 

[MuhammedAbdo]

http://www.guru3d.com/articles_pages/destiny_2_pc_graphics_performance_benchmark_review,5.html
@1440p it's AVG 58 vs 56

??? Are we looking at the same graph? 980Ti is 75fps and FuryX is 58! 980Ti is 30% faster here mate! Are you blind?

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,5.html
@1440p it's AVG 44 vs 50

They test with High Textures, if they tested with Ultra the difference is huge in favor of the 980Ti, here you go:

http://gamegpu.com/action-/-fps-/-tps/middle-earth-shadow-of-war-test-gpu-cpu
http://www.pcgameshardware.de/Mitte...erde-Schatten-des-Krieges-Benchmarks-1240756/
https://www.computerbase.de/2017-10...amm-mittelerde-schatten-des-krieges-2560-1440
https://www.purepc.pl/karty_graficz...th_shadow_of_war_srodziemie_w_ogniu?page=0,10

http://www.guru3d.com/articles_page..._pc_graphics_analysis_benchmark_review,5.html
@1440p memory limited - AVG 40 vs 48

Yeah the difference is relatively smaller in the Guru3D's scenario, but in other places it's hugely in favor of the 980Ti (50% faster and more):

http://www.pcgameshardware.de/Wolfe...us-im-Technik-Test-Benchmarks-Vulkan-1242138/
https://www.computerbase.de/2017-10...nstein-2-1920-1080-anspruchsvolle-testsequenz
http://gamegpu.com/action-/-fps-/-tps/wolfenstein-ii-the-new-colossus-test-gpu-cpu

http://www.guru3d.com/articles_pages/f1_2017_pc_graphics_performance_benchmark_review,6.html
@1440p it's AVG 71 vs 77

Again not testing with max settings, if they did, here is the difference:

https://www.youtube.com/watch?time_continue=330&v=VqABChFrD_w
https://www.computerbase.de/2017-08/f1-2017-pc-benchmark/2/#diagramm-f1-2017-1920-1080

This is choking, definetly.

Yes definitely chocking, check the other 15 games as well, And check the links you post and you will definitely see that! It's horribly bad.

 

[Vya Domus]

They test with High Textures, if they tested with Ultra the difference is huge in favor of the 980Ti, here you go:

Again not testing with max settings, if they did, here is the difference:

Wow , so Fury X is memory starved since it only has 4 GB ?

These findings are truly astonishing. Gimping confirmed.

These discussions are so ridiculous. Why are you people wasting so much of your own damn time to basically prove nothing ?

 

[ikeke]

I'll just say it once again.

If the point is to trip memory limit on FuryX and thus make it faceplant due to lack of memory - you've proven a point that has been acknowledged since the beginning. Lowering a few settings which provide no noteworthy improvements in visuals, though, allows one to play at similar framerate to GTX980Ti.

But I understand, the case here was to prove once and for all that FuryX from 2.5 years ago will faceplant once framebuffer is filled. Good to know. Again.

wrt to Destiny 2 benchmarks, looks like guru3d has updated the results (on 24th November), i copied my data from previous post (16th November) https://www.techpowerup.com/forums/...g-horrible-in-2017.238741/page-4#post-3755209

November 24th, 2017 update - Our Destiny 2 article has been updated. Both AMD and Nvidia now offer a driver that shows massive performance gains in Destiny 2. While AMD's optimized 17.11.2 driver already was included with up to 30% performance increases, Nvidia did a similar thing with their 388.31 drivers. Nvidia also gains over 30% in performance. All graphics cards have been retested, the charts have been updated to reflect this.

 

[MuhammedAbdo]

If the point is to trip memory limit on FuryX and thus make it faceplant due to lack of memory - you've proven a point that has been acknowledged since the beginning. Lowering a few settings which provide no noteworthy improvements in visuals, though, allows one to play at similar framerate to GTX980Ti.

No it's not, most of the games where the card is affected have no VRAM limitation on the FuryX, the card just chocks due to it's bad geometry performance, games like Ghost Recon Wildlands, ME Andromeda, Tekken 7, Agents Of Mayhem, Dishonored Death Of Outsider, Project Cars 2, Forza 7, Total Warhammer 2, Destiny 2, AC Origins, Ark Survival, The Evil Within 2, COD WW2, PUBG, FortNite .. and so on.

 

[Vya Domus]

Geometry shaders are not even used anymore for quite a while since they always affected performance badly. As a matter of fact both Nvidia and AMD ship GPUs with drastically reduced capabilities from this point of view , strangely enough Intel has the most efficient hardware for this. Tessellation replaced this a while ago and even then it isn't that wildly used as there are cheaper methods to fake depth and extra geometry in things like textures.

Goes to show you don't really know what you're talking about and you're just making stuff up as you go along trying to justify your conclusions.

 

[MuhammedAbdo]

Goes to show you don't really know what you're talking about and you're just making stuff up as you go along trying to justify your conclusions.

Geometry performance is both tessellation and polygon rate , this has nothing to do with Geometry shaders! don't know where you are getting this! Your are trying hard and failing!

 

[Vya Domus]

That's just simply incorrect , tessellation and geometry shaders happen at different stages in the pipeline. The geometry stage is a partially serialized process that runs like crap on any GPU inherently.

https://vulkan-tutorial.com/images/vulkan_simplified_pipeline.svg


If you're talking about geometry in a generic fashion then you are massively oversimplifying things to a point where you might just as well not mention it at all.

We get it , you think Fiji is shit , good for you. You don't have to spam this thread with that nonstop and more importantly don't try to bring flawed technical details for your explanations because it makes your points seem even less valid.

 

[cucker tarlson]

And how exactly is new game on fresh drivers for nvidia but game launch drivers for AMD again a valid and final conclusion on cards performance? (or do you really believe that 980Ti is equal to Vega64 LC?)

Context...again. Versus just throwing numbers that paint the picture you like.

Testing in dynamical software environment is possible if you exclude some variables - so taking games that aren't in beta anymore (as most fresh launch titles have been for the last 10y, IMHO) and comparing them on matured drivers that are equally optimized for different uarch's at said games shows the potential of cards, not some optimization that it's [the card] lacking due to driver/game issues.

Based on what you're and other "finewinegar" crowd has been saying we should have buried Ryzen on first week since it's performance on intel optimized scenarios was worse than comparable Intel chip

even more excuses.

If they were gonna improve Fury X performance in drivers, they'd have done that already. They've done that for Vega and Polaris.

Fury X is doing well against 980ti in wolfenstein, but how do you know nvidia isn't gonna release a 30% improvement driver for maxwell next year


Sorry but the point you're making is stupid. It's only true for within the first week or two from launch, and even then it's rare to see a last gen card suddenly regain the lost fps. And it sure as hell happens more often with nvidia optimizing kepler and maxwell for vulkan/dx12 later on than it happens on amd cards.
.

 

[ikeke]

Ok, i get it. AMD = bad. Nvidia = good. Nvidia comes up with magic and AMD never improves. Never has.

Got it. Ok.

 

[MuhammedAbdo]

That's just simply incorrect , tessellation and geometry shaders happen at different stages in the pipeline. The geometry stage is a partially serialized process that runs like crap on any GPU inherently.

You are the one mentioning GS not me, What matters is: FuryX has a very slow polygon rate and tessellation compared to 980Ti, that's what it holds it back in addition to it's limited VRAM, don't try to confuse the matter with irrelevant jargon.

 

[EarthDog]

Weird, how am I NOT involved in a back and forth on this forum?

 

[Vya Domus]

What matters is: FuryX has a very slow polygon rate

No it doesn't , not even close. The fill rate of Fiji is on par with GP104.

and tessellation compared to 980Ti, that's what it holds it back in addition to it's limited VRAM

Although GCN was never good at tessellation none of these new games use it to such an extent that would cripple performance on Fury. Developers aren't retarded.

As a matter of some don't use it at all , like Mankind Divided which employs parallax occlusion mapping which also ironically runs better on Fury compared to the 980ti because it has a higher fill rate.

don't try to confuse the matter with irrelevant jargon.

Of course you think it's irrelevant jargon when all you do is make up pseudo-metrics that have no correlation with the real world.

 

[MuhammedAbdo]

The fill rate of Fiji is on par with GP104.

Sigh, Fill rate is not polygon rate man, educate yourself better next time. Here have a look:

https://techreport.com/review/28513/amd-radeon-r9-fury-x-graphics-card-reviewed/4

Although GCN was never good at tessellation none of these new games use it to such an extent that would cripple performance on Fury. Developers aren't retarded.

Games don't need to use heavy tessellation to cripple the FuryX, they just need to use high amount of polygons in general. That will cripple it faster than 980TI.

 

[Xzibit]

Sigh, Fill rate is not polygon rate man, educate yourself better next time. Here have a look:

https://techreport.com/review/28513/amd-radeon-r9-fury-x-graphics-card-reviewed/4


Games don't need to use heavy tessellation to cripple the FuryX, they just need to use high amount of polygons in general. That will cripple it faster than 980TI.

You do realize in making that argument some of the benchmarks results your holding up contradict it.

 

[Vya Domus]

Games don't need to use heavy tessellation to cripple the FuryX

You can't be for real , you just went from telling us that Fury is crippled by tessellation to saying that actually it doesn't matter after all and that there is yet another obscure unprovable reason behind that.

I guess nothing matters , it's crippled by everything. Isn't that the whole point you're trying to make ? There you have it now stop the nonsense.

Anyway , I'm done. Cool stories you got there what more can I say.

You do realize in making that argument some of the benchmarks results your holding up contradict it.

When you can't keep track of the made up bs you keep coming up with.



It's hard to hide the lack of knowledge while trying to be a fanboy and troll at the same time. It's tough give him a break.

 

[MuhammedAbdo]

You can't be for real , you just went from telling us that Fury is crippled by tessellation to saying that actually it doesn't matter after all and that there is yet another obscure unprovable reason behind that.

Shhh please, stop making a fool out of yourself and showing us how ignorant you are, my statements are clear: Geometry (polygon + tessellation) performance on FuryX is bad, that's why it sucks in most of these games, you are the one confusing Geometry Shaders with fill rate with polygons. Basically your understanding of tech is that of a potato.

It's hard to hide the lack of knowledge while trying to be a fanboy and troll at the same time. It's tough give him a break.

The only fanboy here is the people trying to disprove numbers and benchmarks results with pathetic reasoning and faulty info. Numbers don't lie. But you guys will for sure.

You do realize in making that argument some of the benchmarks results your holding up contradict it.

WOW! Please enlighten us as to how!

 

[xkm1948]

Good debate focus on facts and interpretation of facts.

Bad debate focus on name calling and personal attack.



So who are the good debaters here and who are the actual trolls?