AMD Fury X Reviews

[Silverforce11]

Guys, calm down.

We might actually get to witness some history!

If AMD actually beats Titan X at similar power usage (~250W in gaming loads), I will have to congratulate them. It will be their first time ever with the halo crown and ATI's first in a very long time.

They would have achieved something many deemed impossible. 1) Raw performance victory and 2) Massive efficiency gains.

But on a personal level, I expect more, around 15-20% above Titan X, to make upgrading worthwhile and to make it more playable at 4K on a single GPU on high settings (not maxed).

ps. If only AMD's CPU division is so competitive!!

 

[iiiankiii]

That very same sentiment is used whenever AMD cannot keep up with Nvidia in performance for a given feature.
The very same was said about PhysX. Tesselation.
Getting old.

Sure. Correlate all you want. But, the performance hit is bad. This isn't a AMD vs Nvidia. I'm just stating facts. The performance hit is real. It's up to you to determine if the visual gains are worth it. To me, it isn't.

 

[Skurge]

Sure. Correlate all you want. But, the performance hit is bad. This isn't a AMD vs Nvidia. I'm just stating facts. The performance hit is real. It's up to you to determine if the visual gains are worth it. To me, it isn't.

The gains are obviously worth it... especially when you don't pay for your cards.

I can't remember the last time i used a gameworks feature on my system.

 

[soccerballtux]

So, you suggest:

Gameworks off
Mantle
What about tesselation? PhysX?

Shall we also assume you'd test without AA of any kind if Nvidia handles it without as large a performance hit?

Pretty soon, we are going to go back to ATARI graphics if you turn enough features off to accommodate AMD. Yes, I exaggerate a little but not too far from what is happening.

it was DX11 stuff that started this.

I don't have a problem with tesselation. Just PhysX and Gameworks and dumb wasteful AA techniques nobody wants to use. Weren't you banned a while ago for trying to start stuff like this? Haven't seen you around until recently.

 

[guskline]

I'll try to keep this from evolving into the "Hatfields vs the McCoys" fight (which it sure appears to be shaping up as).

The Fury X appears to have generally caught up to the GTX980TI in most games at stock. TitanX appears to have an edge and it should for its price.

I agree with Keysplayr in this regard. Just test the FuryX, the GTX980TI and the Titan X in how many games you can muster. Do it at 1440p and 4k AND keep all the "Bling" set to the highest level.

Let the chips fall where they may.

I will say this. It looks like AMD finally has a "Big Block" chip that is competitive with Nvidia's TitanX/GTX980TI.

See you at the drag strip!:biggrin:

 

[Aristotelian]

Not exactly a similar comparison. One can only run on the company's card, the other is meant to slow down the competitor's and older generation

I'm not sure that this is an argument to say that x or y should not be used in benchmarks. Back when the PhysX debate was raging on this sub-forum I was one of those that argued that I did not see much of a difference with it on. But I didn't think that PhysX should have been disabled in benchmarks.

I think benchmarks should provide information on the full spectrum of gaming possibilities - on a wide variety of games - and this includes using Mantle etc. because the point of the benchmark should be to show the potential consumer what s/he can do if that card is bought.

 

[Silverforce11]

It looks like AMD finally has a "Big Block" chip that is competitive with Nvidia's TitanX/GTX980TI.

See you at the drag strip!:biggrin:

That's the thing a lot of people are not recognizing the significance of. AMD was the value entity, make smaller GPU that offer good bang for buck but yielded the battle to NV for the performance crown.

People even said they couldn't make a big GPU. The reasons were lower R&D investment, low yields, low returns, no HPC market like NV has to subsidize development of such a monolithic beast...

 

[soccerballtux]

I'm not sure that this is an argument to say that x or y should not be used in benchmarks. Back when the PhysX debate was raging on this sub-forum I was one of those that argued that I did not see much of a difference with it on. But I didn't think that PhysX should have been disabled in benchmarks.

I think benchmarks should provide information on the full spectrum of gaming possibilities - on a wide variety of games - and this includes using Mantle etc. because the point of the benchmark should be to show the potential consumer what s/he can do if that card is bought.

Nah, that's just free advertising for a party abusing the system with a closed technology like PhysX. They don't deserve it. If they opened it and let AMD implement it, I would be fine with that.

The same goes with using a purposely wasteful AA technique simply because it runs less horribly on your architecture-- it's anticompetitive

 

[LTC8K6]

Does anyone think NV will respond once the Fury X performance is known?

Could they have a driver improvement waiting?

It would seem that future driver improvement gains would favor Fury, being the newest chip.

 

[soccerballtux]

Does anyone think NV will respond once the Fury X performance is known?

Could they have a driver improvement waiting?

It would seem that future driver improvement gains would favor Fury, being the newest chip.

They're far enough ahead resources wise I would be surprised if they didn't have this in the back burner just waiting to be turned on. Would be a powerful asset just in case, the opposite of technical debt

 

[Silverforce11]

Does anyone think NV will respond once the Fury X performance is known?

Could they have a driver improvement waiting?

It would seem that future driver improvement gains would favor Fury, being the newest chip.

Do you think Titan X is selling well at $1K when custom 980Ti beats it out of the box and with extra OC headroom?

I wouldn't expect it to be an awesome deal.

As such, what happens to full GM200 dies? They obviously need to move it in volume.. and so, expect a 980Ti Black SKU soon or even Titan X Black. Full GM200, custom designs, 6GB/12GB vram (if its Titan X Black). Basically what Titan X would be if it reached ~1.3 to 1.4ghz out of box boost.

 

[wand3r3r]

Came in for Fury reviews. There sure seems to be a lot of potential.

Left disappointed to see off topic bickering about gameworks, tessellation, physx.

I think the reviews should be in the AMD / NV subforums so we don't have to wade through the bickering unless you want to (in the general forums). Sure we need to compare them (AMD/NV) but it detracts from everything to wade through continuous off topic "debates".

 

[flopper]

It would seem that future driver improvement gains would favor Fury, being the newest chip.

absolutly.
new arc we are in for a treat

 

[railven]

Came in for Fury reviews. There sure seems to be a lot of potential.

Left disappointed to see off topic bickering about gameworks, tessellation, physx.

I think the reviews should be in the AMD / NV subforums so we don't have to wade through the bickering unless you want to (in the general forums). Sure we need to compare them (AMD/NV) but it detracts from everything to wade through continuous off topic "debates".

Agreed. Keep at least the review thread clean, thus put it in respective sub section.

Wanna bicker, open a new thread titled "have at it" or something else in general.

 

[stockwiz]

Would love to see a $550 980 Ti come from this.. that would be cheap enough to get me to bite.... or just buy the AMD card depending on the price.

 

[LTC8K6]

Do you think Titan X is selling well at $1K when custom 980Ti beats it out of the box and with extra OC headroom?

I wouldn't expect it to be an awesome deal.

As such, what happens to full GM200 dies? They obviously need to move it in volume.. and so, expect a 980Ti Black SKU soon or even Titan X Black. Full GM200, custom designs, 6GB/12GB vram (if its Titan X Black). Basically what Titan X would be if it reached ~1.3 to 1.4ghz out of box boost.

I think the belief about more VRAM remains fairly strong, and that belief favors Titan X.

We continue to see cards equipped with more VRAM than they can use, and I think that shows the popularity of more VRAM.

 

[Silverforce11]

https://www.youtube.com/watch?v=YZ3co5a-k1s

OC3D video review is up!

Written review: http://www.overclock3d.net/reviews/gpu_displays/amd_r9_fury_x_review/1

Less power than reference R290X. Similar power to 980Ti!

As usual their charts are all over the place, I don't get the sorting.

 

[zagitta]

So, you suggest:

Gameworks off
Mantle
What about tesselation? PhysX?

Shall we also assume you'd test without AA of any kind if Nvidia handles it without as large a performance hit?

Pretty soon, we are going to go back to ATARI graphics if you turn enough features off to accommodate AMD. Yes, I exaggerate a little but not too far from what is happening.

That's a slippery slope logical fallacy, it's not hard to reach a good middle ground of sane benchmark settings that doesn't favor either brand.

• Gameworks off
• Mantle off
• Tesselation on at sane levels 2-4x for games that don't abuse it *cough* crysis & batman *cough*
• CPU PhysX on, yes the implementation is awful but it's an equal playing field and only hurts AMD because they haven't optimized CPU usage of their driver.
• AA on with non vendor specific types and amount scaling inverse to resolution (AA is barely needed at 1440p and not at all at 4k in my experience)

Back on topic though: There doesn't seem to be any mentions of Fury X on the danish hardware webshops so far so I hope we don't get screwed over


Also NDA should be lifted now!

 

[Silverforce11]

http://www.techpowerup.com/reviews/AMD/R9_Fury_X/

http://www.guru3d.com/articles-pages/amd-radeon-r9-fury-x-review,1.html

http://hardocp.com/article/2015/06/24/amd_radeon_r9_fury_x_video_card_review#.VYqcyfmqpBc

http://www.computerbase.de/2015-06/amd-radeon-r9-fury-x-test/

Close but no cigar! Short of Titan X and trading blows with 980Ti.

Computerbase.de chart

 

[n0x1ous]

and............AT goes live with a motherboard review instead of being on time with Fury X review

EDIT: Apparently Ryan Smith is sick and should have it up in the next few days

 

[BFG10K]

http://www.tomshardware.com/reviews/amd-radeon-r9-fury-x,4196.html

 

[LTC8K6]

So far, it looks like we have a tie with the 980ti.

 

[wolf_squad]

looking at the TPU review, 4096 sp cores, entirely new memory standard, and 2% slower than 980 ti?

disappointing, to say the least...
given amd's history, it's fair to expect some improvements with new drivers, but I hoped they would really be able to deliver the crown home proper this time. alas, no dice.

 

[Borealis7]

So far, it looks like we have a tie with the 980ti.

and they are around the same MSRP...what kind of competition is this?
nVidia's perf/Watt advantage is a big plus in my book (not just for the GTX980TI, but for the entire Maxwell line-up) so things have not changed for me with the release of this card.
still holding on to my GTX680 until 20nm/14nm card come.

 

[LTC8K6]

and they are around the same MSRP...what kind of competition is this?

The type that will probably not do anything for AMD's market share.