FuryX completely abandoned now, barely matching the 1060 or 580!

[MuhammedAbdo]

Borderlands 3: barely at the level of 1060/580 @1080p, 980Ti is 20% faster!
https://gamegpu.com/action-/-fps-/-tps/borderlands-3-test-gpu-cpu

Assassin's Creed Odyssey: slower than a 1060 @1080p, 980Ti is 50% faster!
https://gamegpu.com/action-/-fps-/-tps/assassin-s-creed-odyssey-test-gpu-cpu-2018

Ghost Recon Breakpoint: slower than a 1060 @1080p, 980Ti is 30% faster!
https://gamegpu.com/mmorpg-/-онлайн-игры/ghost-recon-breakpoint-beta-test-gpu-cpu

Metro Exodus: barely at the level of 1060 @1080p at Ultra settings, 980Ti is 25% faster!
https://gamegpu.com/action-/-fps-/-tps/metro-exodus-v-1-0-1-1

Wolfenstein: Youngblood: massively slower than a 1060 @1080p, 980Ti is 90% faster!
https://gamegpu.com/action-/-fps-/-tps/wolfenstein-youngblood-test-gpu-cpu

Crackdown 3: barely faster than 1060 @1080p, 980Ti is 20% faster!
https://gamegpu.com/action-/-fps-/-tps/crackdown-3-test-gpu-cpu

Ace Combat 7: slower than a 1060 @1080p, 980Ti is 40% faster!
https://gamegpu.com/racing-simulators-/-гонки/ace-combat-7-skies-unknown-test-gpu-cpu

Resident Evil 2: barely faster than 1060 @1080p, 980Ti is 70% faster!
https://gamegpu.com/action-/-fps-/-tps/resident-evil-2-test-gpu-cpu

Beyond Two Souls: slower than a 1060 @1080p, 980Ti is 50% faster!
https://gamegpu.com/action-/-fps-/-tps/beyond-two-souls-test-gpu-cpu

The Sinking City: barely faster than 1060 @1080p, 980Ti is 22% faster!
https://gamegpu.com/action-/-fps-/-tps/the-sinking-city-staging-test-gpu-cpu

Draugen: slower than 1060 @1080p, 980Ti is 32% faster!
https://gamegpu.com/rpg/ролевые/draugen-test-gpu-cpu

Total War THREE KINGDOMS: barely faster than 1060 @1080p, 980Ti is 25% faster!
https://gamegpu.com/rts-/-стратегии/total-war-three-kingdoms-test-gpu-cpu

A Plague Tale Innocence: barely faster than 1060 @1080p, 980Ti is 30% faster!
https://gamegpu.com/rpg/ролевые/a-plague-tale-innocence-test-gpu-cpu

Fade to Silence: slightly faster than 1060 @1080p, 980Ti is 28% faster!
https://gamegpu.com/rpg/ролевые/fade-to-silence-test-gpu-cpu

Anno 1800: barely faster than 1060 @1080p, 980Ti is 30% faster!
https://gamegpu.com/rts-/-стратегии/anno1800-test-gpu-cpu

Tropico 6: barely faster than 1060 @1080p, 980Ti is 35% faster!
https://gamegpu.com/rts-/-стратегии/tropico-6-test-gpu-cpu-2

Generation Zero: barely faster than 1060 @1080p, 980Ti is 25% faster!
https://gamegpu.com/mmorpg-/-онлайн-игры/generation-zero-test-gpu-cpu

WRC 8 FIA World Rally Championship: barely faster than a 1060 @1080p, 980Ti is 25% faster!
https://gamegpu.com/racing-simulators-/-гонки/wrc-8-fia-world-rally-championship-test-gpu-cpu

Age of Wonders Planetfall: slower than a 1060 @1080p, 980Ti is 22% faster!
https://gamegpu.com/rts-/-стратегии/age-of-wonders-planetfall-test-gpu-cpu

Ancestors: The Humankind Odyssey: barely faster than a 1060 @1080p, 980Ti is 22% faster!
https://gamegpu.com/rpg/ролевые/ancestors-the-humankind-odyssey

Dirt Ralley 2: barely faster than a 480 @1080p, 980Ti is 25% faster!
https://gamegpu.com/racing-simulators-/-гонки/dirt-rally-2-0-test-gpu-cpu

Assetto Corsa Competizione: slower than a 1060 @1080p, 980Ti is 60% faster!
https://gamegpu.com/racing-simulators-/-гонки/assetto-corsa-competizione-0-5-2-test-gpu-cpu

 

[Solaris17]

Isn't the 980ti still a monster?

 

[MuhammedAbdo]

Isn't the 980ti still a monster?

Yeah it is, the problem is, the FuryX has fallen so far behind it, it's now barely at the 1060/580 level of performance in the vast majority of new titles, this has been happening since 2017.
The FuryX used to be just 10-15% slower than a 980Ti.

 

[natr0n]

980ti user.

Sometimes it seems like this card is ridiculously overpowered.

 

[oxrufiioxo]

Fine wine....

 

[xkm1948]

Oof, that is brutal

Welp I no longer have mine so doesn't really care any more. Meanwhile enjoy the red death squad coming in here accusing you of "Meh AMD can do no wrong" or "You are an idiot of buying a 4GB card in the first place"

Have fun

 

[eidairaman1]

Oof, that is brutal

Welp I no longer have mine so doesn't really care any more. Meanwhile enjoy the red death squad coming in here accusing you of "Meh AMD can do no wrong" or "You are an idiot of buying a 4GB card in the first place"

Have fun

What gets me is you think nvidia is innocent when they are far from it

 

[xkm1948]

What gets me is you think nvidia is innocent when they are far from it

Now quote me exactly where I said or implied "NVIDIA is innocent"

You can't, because I did not say or imply such a thing.

Liar

 

[chr0nos]

Well it has 96 ROPS vs 64 ROPS on the FuryX and 6Gb instead of 4Gb.

Theres only so much driver optimizations they can do.

 

[Keullo-e]

Weird, since no problems with R9 290 cards here even with newer games. And when a game supports Crossfire (which is pretty rare in newer games..), VRAM is the worst bottleneck.

Though I rather check benchmarks from somewhere else than some neverheard Russian site.

 

[FordGT90Concept]

The problem is 4 GiB VRAM. Even at 1920x1080, most games exceed that which means the card is having to wait for much slower RAM to service it.

 

[Outback Bronze]

In all fairness the Fury X was designed to be a 4K card.

Lets see these results at 4K then make an informed decision yeah?

 

[blobster21]

In all fairness the Fury X was designed to be a 4K card.

Lets see these results at 4K then make an informed decision yeah?

but then

The problem is 4 GiB VRAM. Even at 1920x1080, most games exceed that which means the card is having to wait for much slower RAM to service it.

In all fairness, this card is in the same boat as the 9700Pro and the GTX295 : things of the past

 

[newtekie1]

The problem is 4 GiB VRAM. Even at 1920x1080, most games exceed that which means the card is having to wait for much slower RAM to service it.

That isn't really true. Most games might be reserving over 4GB of RAM at 1080p, but they aren't actually using it. And if they are using it, they're just cramming a bunch of unused textures in video memory, so it doesn't really affect performance to only have 4GB. There are a few excpetions to this, but for the most part games don't need more than 4GB@1080p.

Yeah it is, the problem is, the FuryX has fallen so far behind it, it's now barely at the 1060/580 level of performance in the vast majority of new titles, this has been happening since 2017.
The FuryX used to be just 10-15% slower than a 980Ti.

When the 1060 launched, the Fury X was less than 10% faster than the 1060 at 1080p. And when the 580 launched the Fury X was right at 10% faster at 1080p. And the fact was the architecture that the Fury X was based on had pretty much already received all the driver optimizations AMD could do. So it wasn't going to get any better with age. But the 1060 and 580 were both newer architectures, that could get benefits from more driver optimizations. It isn't that it was abandoned, it was just already optimized pretty much as far as it could go when it was released. And it already kind of was at 1060/580 levels of performance.

 

[oxrufiioxo]

What gets me is you think nvidia is innocent when they are far from it

I agree, the 680 started struggling in games only a couple years after it launched because Nvidia cheaped out and went with a 2GB framebuffer.

 

[biffzinker]

At least the Radeon VII isn't short on capacity with 16GB.

 

[Valantar]

Hm. While the Fury X was never a good 1080p GPU, this is a bit disappointing. I guess I've been right to hold off on buying any of the newer titles I'm interested in (Wolfenstein, Control, Borderlands, a few others) until the faster Navi GPUs start arriving. While the longevity of the 980Ti is very impressive, I still can't say I'm disappointed in how long my Fury X has kept me happy though. 4 years of games at 1440p? That's not bad at all, even with some IQ reductions in later years.


Edit: Hm. GameGPU has a weird way of drawing bar charts. Looking at the Metro Exodus results in 1440p Medium, the Fury X does 24/49 FPS (presumably 99% or min./avg.) while the 980Ti does 35/45, yet the 45 bar of the latter is noticeably longer than the 49 bar of the former. Do these graphs use the bar of the minimum/99% as the zero point for the average fps bar? While I guess you could argue that this serves to take focus away from only average FPS (which is obviously good), it's a very confusing and borderline misleading way of drawing graphs. Not that this affects what the OP is reporting whatsoever, it's just a very curious way of visualizing data.

 

[Frick]

Soo ... abandoning it? Do you mean if they tweaked the drivers it would be faster? I really don't get the point. Yeah it's slow.

Oof, that is brutal

Welp I no longer have mine so doesn't really care any more. Meanwhile enjoy the red death squad coming in here accusing you of "Meh AMD can do no wrong" or "You are an idiot of buying a 4GB card in the first place"

Have fun

Yeah, you are having fun.

 

[Peter Lindgren]

Still rocking my Nano.

Replaced fan with a Noctua. It never goes over 2000rpm and is really quiet. Test below is with WOT @1440p with Ultra settings. FPS are between 60-80. With a Freesync monitor it runs very smooth. Mem goes to the max but no hickups yet. No need to upgrade yet.

 

[jesdals]

Well at least you had a good run for it - being a Radeon VII owner, thats a hard turn - but did now when I got it that Navi would be the end of Vega

 

[Vya Domus]

I would have thought more recent examples with baffling results would be more interesting compared to ancient history.

A 6GB 1060 outperformed by a a 4GB RX570, even at 4k ? I guess not, AMD has a way to spark interest even long after their products become obsolete. I must say though, a 4 year old card not being at the top of the heap anymore is indeed shocking, even though there were instances of the 980ti being 30% faster even at launch. But I digress, AMD bad.

Oh, and gamegpu, everyone's go to benchmarking site, this is definitely not the first time I am hearing about it. Nice.

In all fairness the Fury X was designed to be a 4K card.

Yes, 4 years ago. Times change, the framebuffer remains.

 

[londiste]

Edit: Hm. GameGPU has a weird way of drawing bar charts. Looking at the Metro Exodus results in 1440p Medium, the Fury X does 24/49 FPS (presumably 99% or min./avg.) while the 980Ti does 35/45, yet the 45 bar of the latter is noticeably longer than the 49 bar of the former. Do these graphs use the bar of the minimum/99% as the zero point for the average fps bar? While I guess you could argue that this serves to take focus away from only average FPS (which is obviously good), it's a very confusing and borderline misleading way of drawing graphs. Not that this affects what the OP is reporting whatsoever, it's just a very curious way of visualizing data.

They do use minimum as the zero point for the average FPS bar.
This actually is not a bad approach, I just wish they would be clear more about it.

Having tried drawing overlapping charts with a couple different solutions, their approach is probably motivated by it being very simple

 

[Sybaris_Caesar]

I don't know why everyone's shooting the messenger. Gamegpu has been quite popular for the last few years cause of their huge list of GPUs and CPUs benched. It was called gamegpu.ru. Then increased demand made them buy a bigger server/bandwidth (idk the shenanigans alright?) a few years back iirc.

 

[64K]

Isn't the 980ti still a monster?

I have a 980 Ti and it has seen it's glory days but it's no monster any longer. I am in need of an upgrade due to gaming at 1440p but I'm debating waiting on the 3XXX series or maybe pick up a 2070 Super which is around the same level of performance as a 1080 Ti.

My 980 Ti is soundly beaten even by a 2060 FE which is around 35% faster than my 980 Ti.

 

[delshay]

@MuhammedAbdo

I'm looking into this as I also have lower score with Superposition with latest drivers. I will be going back to Crimson drivers to see if anything has changed.

With the latest drivers it does not seem to be holding utilization very well, but will confirm if anything has changed between Crimson & Adrenalin drivers. Something for sure is holding my card back, but it could also be my modded hardware on the R9 Nano.