Official Fury X specs

[Fire&Blood]

Reads: Content under embargo until ?AM June 24, 2015 ???

So June 24 it is for Fury/X, and the 18th for the 300 series. Reviews on Fury should be interesting. Insane numbers if it ends up like that on reviews, must justify them having a 5k monitor there at the presentation.

The slide is from a AMD presentation in China that just finished a few hours ago, it's here: http://live.pconline.com.cn/483.html

 

[RampantAndroid]

The benefit of using driver level optimization means that potentially every GCN card could get an effective memory size boost.

In one of the other dead AMD speculation threads, I copy/pasted a bunch from the DirectX spec - memory management isn't up to the driver, it's a relationship between DirectX and the rendering engine. DX will remove items not used for a while, unless they're tagged as being required (at which point they take priority over other items.) The documentation emphasizes that it's up to you, the developer, to manage memory best. As such, I don't see how AMD can do anything....short of just removing random stuff from memory.

Granted, I didn't read the DX12 spec, but even if DX12 comes with the improvement, many games already released (and many more due to be released in the relatively new future) will not be on DX12.

 

[96Firebird]

Changes everything. I was convinced I absolutely needed 2 cards for 4K and was still concerned about compromising settings/frame rate

Ah, the good 'ol asterisk. It's always a wonder what is behind that...

 

[Silverforce11]

For a comparison to TMU/ROP performance, Texture & Pixel fill rate:

This is 285/Tonga ROPs: Note 32 Tonga ROPs is faster than 64 Hawaii ROPs

~R295X2 Texture Fill and ~4x R290X Pixel Fill.

 

[SimianR]

Weird - I guess all the leaked specs were wrong about 128 ROP's then? I'm surprised that they are left entirely untouched at 64 - the same amount as the R9 290x. Perhaps the rumors were referencing the dual GPU fury.

 

[Silverforce11]

Weird - I guess all the leaked specs were wrong about 128 ROP's then? I'm surprised that they are left entirely untouched at 64 - the same amount as the R9 290x. Perhaps the rumors were referencing the dual GPU fury.

Untouched?

Look at the Pixel Fill-rate.

 

[Eymar]

Untouched?

Look at the Pixel Fill-rate.

The pixel fill-rate in slide matches values listed in 290x slides (http://www.legitreviews.com/wp-content/uploads/2014/04/r9-specs-645x498.png). As been stated before pixel fill-rate shouldn't be an issue even very high res scenarios (3x 1440p, 4k, etc.). Have to save transistors somewhere and if more pixel fill-rate doesn't get you anything, I'd say wise decision not to waste silicon on it. Though, please correct me if I'm wrong as I too was expecting ROP increase as been the norm with each new GPU arch.

 

[imported_bman]

Do not know how the systems compare though.

 

[Erenhardt]

^ we don't know which part of the game was tested. You can't compare results from different scenes.

The rest of the system have close to no impact on the performance, granted both use a decent high-end hardware.

 

[LTC8K6]

Why is there only one score for the the CF XT?

 

[Cookie Monster]

Do not know how the systems compare though.

Finally some performance numbers. Although its not apples to apples, those are some promising numbers!

 

[LTC8K6]

Finally some performance numbers. Although its not apples to apples, those are some promising numbers!

If only we knew what the asterisk meant?

 

[imported_bman]

^ we don't know which part of the game was tested. You can't compare results from different scenes.

The rest of the system have close to no impact on the performance, granted both use a decent high-end hardware.

My bad, was thinking that Far Cry 4 had a bench run.

 

[LTC8K6]

Might depend on drivers and actual settings?

 

[RussianSensation]

The pixel fill-rate in slide matches values listed in 290x slides (http://www.legitreviews.com/wp-content/uploads/2014/04/r9-specs-645x498.png). As been stated before pixel fill-rate shouldn't be an issue even very high res scenarios (3x 1440p, 4k, etc.). Have to save transistors somewhere and if more pixel fill-rate doesn't get you anything, I'd say wise decision not to waste silicon on it. Though, please correct me if I'm wrong as I too was expecting ROP increase as been the norm with each new GPU arch.

It is not correct to compare theoretical pixel fill-rate on paper, not across AMD vs. AMD or across AMD vs. NV. Every time this point is brought up, it's either brushed aside or purposely ignored. Like some gamers tried to compare GM200's 96 ROPs to AMD's 128 ROPs and use GPU clock speeds to support their position that AMD will need 128 ROPs clocked at 1-1.05Ghz to match GM200's 96 ROPs clocked at 1.2Ghz Boost. That's not how it works at all. Different GPU architectures utilize ROPs differently. Just because 2 GPUs have the same number of ROPs, it doesn't tell us the actual real world throughput. We can only guesstimate and often we will be wrong.

"Even with the same number of ROPs and a similar theoretical performance limit (29.6 vs 28.16), 7970 is pushing 51% more pixels than 6970 is." ~ AT

We have already seen 285 outperform a 290 in pixel-fill rate tests. Even though these tests aren't 100% pixel fill-rate limited since they are also memory bandwidth dependent, the point is you can have a card with 32 ROPs that puts down the same or higher pixel fill-rate than a 64 ROP card. For that reason we can't just compare Hawaii's 64 ROPs to Fury's 64 ROPs. I bet it's not that simple at all. It's not out of the question that AMD's pixel fill-rate in the real world has increased 70-100% with the same number of ROPs. However, based on theoretical differences between Hawaii's and Tahiti's ROP throughput, Hawaii's ROPs were highly inefficient. We need to wait for benches but I am confident Fury's pixel fill-rate performance will trounce 290X's by a lot.

As such, I don't see how AMD can do anything....short of just removing random stuff from memory.

Unless you have a 4K monitor and plan on buying 2 $550+ cards and keeping them > 3 years, this is probably a non-issue. 980 SLI beats Titan X at 4K in almost all games where SLI scales, including GTA V. The chance that a single Fury X will become VRAM limited before it becomes GPU limited is very slim. We really need to wait for benchmarks and overclocking results because these matter a lot more than 4GB vs. 6GB of VRAM. A lot of gamers who buy a pair of $550+ cards will be reselling them and going with 14nm/16nm HBM2 cards as I don't think too many elite PC gamers spend $1100-1300 on GPUs and keep them for 5 years. It's just not a very good upgrading strategy.

It's also game dependent. If someone is planning to get a $200 racing wheel and play Project CARS and can't live without PhysX in Batman (say this gamer's favourite franchise), well the 980Ti wins automatically. Similarly someone might choose 2x Fury cards since they don't want 600W+ of power exhausted into their case. Also, knowing NV's driver support for Fermi and Kepler, it's a tall order to bet that with time NV's card will age better. History is highly against nV in this regard as time has shown that AMD's GCN cards age a lot better than Fermi and Kepler. What guarantee is there that NV will spend as much time optimizing for Maxwell once they release Pascal next year? We know AMD has no choice but to optimize GCN since for 2016 they still plan to use GCN as per their Financial Analyst Day 2015 slides. From a driver perspective, that makes Fury a 'safer' bet unless one plays GW titles predominantly.

 

[Grooveriding]

4:53 of the Computex presentation video: https://youtu.be/QQ92qWdVLsM


Now if this is true:

Changes everything. I was convinced I absolutely needed 2 cards for 4K and was still concerned about compromising settings/frame rate

That is crazy performance if accurate. I hope it holds up in reviews.

 

[TemjinGold]

So I just realized something. The Dual Fiji card has 2x 8-pin. Fury XT also has 2x 8-pin. Holy crap that's going to be a LOT of OC headroom....

 

[flopper]

So I just realized something. The Dual Fiji card has 2x 8-pin. Fury XT also has 2x 8-pin. Holy crap that's going to be a LOT of OC headroom....

and 2xnano cards and I can keep my PSU without upgrading, amazing

 

[CakeMonster]

Can we please set the record straight on what memory compression does? People are claiming all sorts of things everywhere I've been reading about Fury today.

My impression was that compression can drastically increase effective memory bandwidth (980 was a prime example), but it doesn't reduce actual memory usage in Mbs, since the work needs to be done on the decompressed data in memory.

 

[happy medium]

[
wrong thread.

 

[RoarTiger]

edit

 

[AnandThenMan]

Please report those posts and not respond to them. The splitting into AMD/Nvidia sections was a really great idea IMO but some of the nonsense is still seeping in.

...BTW I was highly skeptical that Fury would turn out to the official name, guess I didn't have much faith in AMD doing something good on a marketing level.

 

[happy medium]

What exactly do Titan SLI X benchmarks have to do with the OP or the post you quoted? Can we not shit up this thread with useless graphs? Some people may be looking for actual Fury X information.

Actually I apologize, its this new thread system , wrong thread.

 

[IEC]

edit

Just report the troll, don't fall to his level.

 

[xthetenth]

Can we please set the record straight on what memory compression does? People are claiming all sorts of things everywhere I've been reading about Fury today.

My impression was that compression can drastically increase effective memory bandwidth (980 was a prime example), but it doesn't reduce actual memory usage in Mbs, since the work needs to be done on the decompressed data in memory.

The answer is pretty much "who knows?" because compared to previous uses of compression, this use is specifically mentioned as being for the purpose of making more out of less memory, which means it likely is reducing usage, and that hasn't been seen before.