AMD's Fury and the future of the enthusiast GPU

The thing i dont get is that aiming for 4k gaming has been a thing for quite a while now, but every benchmark i see has fps averaging around like 40-50. Considering 4k monitor and a 4k card each cost an arm and a leg, i wonder whats the point of either playing a slideshow or on potato settings.

 

I find 40-50 fps perfectly fine. Besides that, one can usually go from 'ultra' settings down to high and get a significant fps increase without noticing any worse graphics.

Anandtech writes that the Fury nano will be a full Fury without any cuts, that is indeed interesting. But the Fury with air cooling will be cut down.

http://www.anandtech.com/show/9390/the-amd-radeon-r9-fury-x-review/10


Edit: Also, they mention an interesting reason why the Fury X is slower in 1440p and lower resolutions: The inefficient AMD drivers that tax the CPU more. I would have never thought of that, but it makes sense.

 

But youre getting 40 fps on top tier card with games that have come out in last 1-2 years. What are you going to get in a year? 20?
Future proofing is foolish, but this seems like it will loose value ridiculously fast.

 

I don't think we will see that much higher requirements in the next year, but yes high-end graphic cards always lose their value fast, just look at the Nvidia Titan (predecessor to the Titan X). I'm not recommending anyone getting a 700€ graphics card.

 

AnandTech: "The AMD Radeon R9 Fury Review, Feat. Sapphire & ASUS."

 

Seems like a good card that wasn't affected much by the reduced amount of shaders. It costs 10% more than a 980 and is 10% faster than a 980 so no surprises there. But the full Fury X is only 100$ more and you can get a 980Ti for that price as well. Not sure what to think about this card to be honest.

 

Bottom line: Both Fury cards need to be 50$ cheaper to be competitive.

 

That makes a lot of sense.

 

Look at the performance of AMD cards in a CPU limited situation:
http://uk.hardware.info/reviews/614...th-an-amd-cpu-testnresultsn--1920x1080-medium
Yup their drivers are that bad.

 

Overclocker's dream my ass, Furys overclocking headroom is mediocre: https://www.techpowerup.com/reviews/AMD/R9_Fury_X_Overvoltage/

 

AMD is providing performance and perf/W results of the R9 Nano against the R9 Fury X and the R9 290X in Far Cry 4, if these slides are legitimate.



From Golem.de.

The performance ratio between the 290X and the Fury X in AMD's slide is basically the same as in TechPowerUp's Far Cry 4 performance results.

Code:

AMD

R9 Fury X  42.1 FPS  100%
R9 Nano    33.4 FPS   79%
R9 290X    30.5 FPS   72%


TechPowerUp (3840x2160)

R9 Fury X  39.5 FPS  100%
R9 390X    31.9 FPS   81%
R9 290X    28.1 FPS   71%
GTX 980    27.8 FPS   70%

I can't extrapolate from just one game, but given what I expect from the Nano, I wouldn't be surprised if it was around 390X/980 levels of performance in general.

 

I just read the entire thread because I was bored. A little bit plug, lititle bit off topic, whole lot of necro, but star citizen runs fine on SLI triple monitor setup @1080p with just a single 970 after the recode. If you are having problems with your current setup, I'd check against 970 benchmarks for your next purchase.

On topic, fury isn't worth it. Essentially unless you plan to be over clocking like a boss (which isn't really worth it), the per watt metric doesn't help you much unless performance is much closers than that. This is not an intel K and you won't make up the difference if you are looking for the best performer.