Review Index:
Feedback

Sapphire Radeon R9 Fury 4GB Review with CrossFire Results

Author:
Manufacturer: Sapphire

Fiji brings the (non-X) Fury

Last month was a big one for AMD. At E3 the company hosted its own press conference to announce the Radeon R9 300-series of graphics as well as the new family of products based on the Fiji GPU. It started with the Fury X, a flagship $650 graphics card with an integrated water cooler that was well received.  It wasn't perfect by any means, but it was a necessary move for AMD to compete with NVIDIA on the high end of the discrete graphics market.

View Full Size

At the event AMD also talked about the Radeon R9 Fury (without the X) as the version of Fiji that would be taken by board partners to add custom coolers and even PCB designs. (They also talked about the R9 Nano and a dual-GPU version of Fiji, but nothing new is available on those products yet.) The Fury, priced $100 lower than the Fury X at $549, is going back to a more classic GPU design. There is no "reference" product though, so cooler and PCB designs are going to vary from card to card. We already have two different cards in our hands that differ dramatically from one another.

The Fury cuts down the Fiji GPU a bit with fewer stream processors and texture units, but keeps most other specs the same. This includes the 4GB of HBM (high bandwidth memory), 64 ROP count and even the TDP / board power. Performance is great and it creates an interesting comparison between itself and the GeForce GTX 980 cards on the market. Let's dive into this review!

Continue reading our review of the Sapphire Radeon R9 Fury 4GB with CrossFire Results!

Every new graphics card review should start with a little analysis of the GPU itself. Obviously the architecture improvements and capabilities of the cut-down Fiji GPU on the AMD Fury are identical to those found in the Fury X, so if you need that information I would ask that you check back into our story from last month that go over those changes in great detail.

  R9 Fury R9 Fury X GTX 980 Ti TITAN X GTX 980 TITAN Black R9 290X
GPU Fiji Pro Fiji XT GM200 GM200 GM204 GK110 Hawaii XT
GPU Cores 3584 4096 2816 3072 2048 2880 2816
Rated Clock 1000 MHz 1050 MHz 1000 MHz 1000 MHz 1126 MHz 889 MHz 1000 MHz
Texture Units 224 256 176 192 128 240 176
ROP Units 64 64 96 96 64 48 64
Memory 4GB 4GB 6GB 12GB 4GB 6GB 4GB
Memory Clock 500 MHz 500 MHz 7000 MHz 7000 MHz 7000 MHz 7000 MHz 5000 MHz
Memory Interface 4096-bit (HBM) 4096-bit (HBM) 384-bit 384-bit 256-bit 384-bit 512-bit
Memory Bandwidth 512 GB/s 512 GB/s 336 GB/s 336 GB/s 224 GB/s 336 GB/s 320 GB/s
TDP 275 watts 275 watts 250 watts 250 watts 165 watts 250 watts 290 watts
Peak Compute 7.20 TFLOPS 8.60 TFLOPS 5.63 TFLOPS 6.14 TFLOPS 4.61 TFLOPS 5.1 TFLOPS 5.63 TFLOPS
Transistor Count 8.9B 8.9B 8.0B 8.0B 5.2B 7.1B 6.2B
Process Tech 28nm 28nm 28nm 28nm 28nm 28nm 28nm
MSRP (current) $549 $649 $649 $999 $499 $999 $329

The Fiji Pro GPU differs from the Fiji XT found on the original Fury X in a couple of important ways. First and foremost, the stream processor count drops from 4096 down to 3584, a loss of 512 or 14%. That means AMD's Fury card will include 56 compute units compared to the 64 compute units found in the Fury X and along with the stream processor change comes a decrease in associated texture unit count as well. That drops from 256 to 224, the same 14%.

The "up to" rated clock speed on the Fury GPU is 1000 MHz though we already know of some overclocked models that are shipping on day of release (which is July 16th). Those overclocks appear to be minimal (~40-50 MHz) so don't expect a lot of variance quite yet. That 1000 MHz reference speed is just 50 MHz shy of where the Fury X ran with the water cooler integration; I know that some users will be surprised that there aren't more performance gains and headroom found in that original Fury X product.

View Full Size

The memory sub-system remains unchanged on the Fury from the Fury X. We are still using the stacked memory with interposer combination dubbed HBM, high bandwidth memory. This is another pretty complex topic that if you are unfamiliar with is worth reading up on. We have an editorial on the dramatic changes that HBM offers for GPUs that you should comb over. The memory clock speed of just 500 MHz remains over the 4096-bit memory interface, resulting in 512 GB/s of total memory bandwidth, the same as on the Fury X. Also, there are still 64 ROPs / raster operators on Fury, the same number that ships on Fury X.

Interestingly, the TDP of the Fury is identical to that of the Fury X, despite that fact that we are running at a slightly lower clock speed and with 14% fewer stream processors in operation. We are likely seeing the result of AMD's Fiji GPU binning to get the best GPUs on the flagship Fury X product. That improved silicon will often run at lower voltage thresholds for the same performance level and thus the "slightly worse" GPUs that make their way into the retail Fury cards need a bit more juice to get the 3,584 SPs up and running stable. In our real world testing this is echoed as well, where the Sapphire R9 Fury card uses just 6-7 watts less power than the Fury X sample.

Peak theoretical performance drops from 8.6 TFLOPS down to 7.2 TFLOPS with the decrease in clock speed and shader count, but that is still a number that exceeds the highest rated GM200 GPU. But as we know from our previous Fury X review, TFLOPS ratings do not correlate to actual gaming performance all the time.

So now we know what is the same - what is different? It's all in the implementation.

Video News


July 10, 2015 | 11:14 AM - Posted by BDK (not verified)

Sapphire always make sexy looking cards.

July 10, 2015 | 11:18 AM - Posted by Mobile_Dom

these are better than i expected, but at this point im just waiting to see price and performance of the Nano, with a Fiji XT core on something that tiny and using only 175w-ish, thats gotta be clocked super low

July 10, 2015 | 11:32 AM - Posted by Anonymous (not verified)

Crossfire performance is massive. Nvidia stands no chance with their SLI in many games and the others are close.

July 10, 2015 | 12:28 PM - Posted by obababoy

I agree that the performance is better in CF, but the problem is a majority of people only want ONE card. With that said this does great as a single card.

~AMD fan.

July 10, 2015 | 05:49 PM - Posted by Anonymous (not verified)

When the Fury X2 comes out I'm sold.

July 10, 2015 | 11:40 AM - Posted by Edmond (not verified)

Price/performance is still SHIT.

I see no point in buying anything besides a 970 or 290, tbh.
or using two ... if you really think you need it.

July 10, 2015 | 11:46 AM - Posted by Dusty

Compared to the 980, a 10% increase in price, for a 5%-30% increase in performance is not shit. If you think there is no point for these cards, then you obviously don't have to buy one. I am debating getting a Fury to replace my 7950 for 1440p gaming.

July 10, 2015 | 11:48 AM - Posted by Edmond (not verified)

get two 290s/970s for the same price

July 10, 2015 | 12:04 PM - Posted by Dusty

I can't be arsed to mess with Crossfire/SLI, and my case would be a cramped nightmare with two full size cards.

Two 970s are at least $600, and two 290s are $540, so only $10 off of MSRP of the Fury. Do you have any benchmarks to compare, or should I go and find that information too?

July 10, 2015 | 12:32 PM - Posted by obababoy

Two 290's would spank this card but you are right the complications with power and CF are not worth it to me. I'd rather pay the premium for a top end single GPU.

July 10, 2015 | 12:50 PM - Posted by Mordorson (not verified)

Could be a pretty good option with frame rate control being available in the new 5.7 drivers. Couple with a Freesync monitor and you've got butter smooth gaming for the next 2 years.

July 10, 2015 | 12:54 PM - Posted by obababoy

That is an interesting prospect. the hawaii GPU's get oddly efficient when downclocked to 900mhz. So running CF with my MG279Q could yield a very nice upgrade for a couple years at around $240 for the 290.

July 10, 2015 | 11:48 AM - Posted by Anonymous (not verified)

Any chance to see the GTA 5 bench marks?

July 10, 2015 | 11:55 AM - Posted by Dusty

"NVIDIA still has a fighting chance with the GeForce GTX 980 of course: it is $50 less expensive, includes a free game (currently at least) and has the advantage of more frequent and usually more reliable drivers behind it. But AMD has stepped up its game and released a high-end GPU that should make NVIDIA worry. And that's great for everyone."

While I agree that Nvidia has had more frequent driver releases, I don't know if the claim that their drivers are more stable can be substantiated.

July 10, 2015 | 12:33 PM - Posted by Doeboy (not verified)

AMD drivers will also improve with time while Nvidia drivers are not likely to squeeze out better performance since it's been out for several months.

July 10, 2015 | 03:20 PM - Posted by Anonymous (not verified)

AMD fans always say this and history has shown it just really doesn't happen.

July 10, 2015 | 04:28 PM - Posted by Anonymous (not verified)

That's false. GCN used to be lower than Kepler. Right now, GCN has overtaken Kepler in performance. History has shown that like with all new GPU archs, performance does increase with time as driver matures.

December 5, 2015 | 09:14 PM - Posted by Eric Magana (not verified)

Well, for now that has changed. The fury dropped to 499 in price on newegg (in the US) and it includes star wars battlefront.

July 10, 2015 | 12:50 PM - Posted by Hani

fix the jump link to the benchmarks page in page# 3, it is directing to AMD's 7990 benchmarks!

July 10, 2015 | 01:08 PM - Posted by obababoy

Great card but not worth an upgrade from my Sapphire Vapor-X R9 290. I might have to wait for the next iteration...That and the MG279Q I just bought extends longevity even more! 1440p max settings is pretty attainable at 35fps with the 290.

July 10, 2015 | 01:06 PM - Posted by Shortwave (not verified)

Well, due to the fact you guys don't benchmark using resolutions that 90% of people still use.. Your benchmarks just got much less useful to me.

July 10, 2015 | 01:10 PM - Posted by obababoy

Strangely Anandtech did the same thing. Maybe because they were rushing to get it out first? I am finally on a 1440p monitor so I am satisfied, but overall Fury X struggled at 1080p compared to 980ti so I would be curious how this does.

July 10, 2015 | 04:05 PM - Posted by Anonymous (not verified)

I am hesistant to boost other sites, but since PCPER has no 1080p testing anymore I can say that Sweclockers do.

And on 1080p, this card gets beaten by 980 by a few percentage points consistently, and that's without any OC'ing. And we all know about Maxwell's massive overclocking capability.

So at 1440p or above, the Fury is a great card. At 1080p it gets beaten and as you said, that is where most people still are (and will remain for the foreseeable future). Maxing out Witcher 3 on ultra at 1080p means that even the 980 Ti dips below 60 fps.

And there still is no card that is 4K-ready. Look at GTA V. Even Big Pascal probably won't breach 60 fps on ultra at 4K on GTA V.

July 10, 2015 | 05:08 PM - Posted by Heavy (not verified)

why would they include 1080p benchmarks this cards arent meant for 1080p player it just wouldnt make sense to have 500+ card and 100 dollar monitor.i think pcper is targeting the right audince with this benchamarks

July 10, 2015 | 05:47 PM - Posted by Anonymous (not verified)

1080 is important for the 120\144hz gamers still.

July 10, 2015 | 11:12 PM - Posted by doeboy (not verified)

How is it important when it can play all games smoothly at 1080p? Same with the GTX 980. It's pointless who wins at that resolution when both can do the job.

July 15, 2015 | 02:56 PM - Posted by Anonymous (not verified)

"Smoothly" isn't 60 fps. It's 144 fps. If you disagree your eyes either suck or you have never tried true 144 fps gaming.

P.S. Witcher 3 at ultra on 1080p can't even hit 60 fps on the mighty 980 Ti on most benchmarks I've seen, so forget hitting 100+ fps.

If AMD isn't 1080p on its high-end cards its not relevant at all.

July 23, 2015 | 05:00 AM - Posted by Anonymous (not verified)

Asynchronous monitors like GSYNC are smoother at lower frame rates so keep that in mind for high-end gaming.

Aiming for 144FPS with good quality settings is always going to be problematic... cards will get better but games will also get more demanding.

For 1080p benchmarks use Techpowerup, though I don't quite agree with the scores since you should use less AA at higher resolutions to have a more apples-to-apples comparision of visuals. Roughly speaking 30 to 40% higher frame rates is what I'd estimate (i.e. same settings except 8xMSAA 1080p vs 4xMSAA 1440).
http://www.techpowerup.com/reviews/MSI/GTX_980_Ti_Gaming/24.html

July 11, 2015 | 02:54 PM - Posted by Simjedi (not verified)

Agreed, 1080p is still very relevant and it's sad they don't test it anymore.

For those saying high end cards like this aren't meant for 1080p I couldn't disagree any more. If you like to crank up the eye candy with loads of AA 1080p testing is relevant.

Before anyone starts saying "get a higher resolution panel" a lot of people game in their living room. I have a 1080p projector at 130" and will never go back to a tiny monitor. And don't even suggest a 4k one as they are high $$.

July 27, 2015 | 07:05 PM - Posted by manxusco (not verified)

Even My 980 sc can be taxed in some high end games at 1080p oh, by the way, my 1080p monitor cost a good bit more than $100.00.
1080p is still relevant for me.

July 10, 2015 | 01:15 PM - Posted by Prodeous

Like the performance, don't like the price. And for the cooling solution, there is so much exhaust space at the back, yet this cooler uses nearly next to none of it :(

Do wonder who will release a water cooled version or more blower version...

July 10, 2015 | 01:25 PM - Posted by obababoy

I agree with the price being too high for me, but exhaust space at the back...Barely any heat ever comes out of these vents on an aftermarket GPU like this. Especially considering most use vertically aligned heatsink fins. I think keeping the shorter PCB vs what Asus did was a smart move because the fan blows the air right through the heatsink and doesn't get trapped anywhere but in the case.

July 10, 2015 | 03:23 PM - Posted by Anonymous (not verified)

I thought the big selling point with these Fury cards was they are so small to fit in small cases. So to have to extend the PCB and/or cooling takes away that advantage and just shows are poor heat/power is on AMD cards.

July 10, 2015 | 03:19 PM - Posted by Anonymous (not verified)

The Sapphire card looks really dumb with the extended cooling from the PCB.

July 15, 2015 | 08:04 AM - Posted by obababoy

I think it is a great idea. straight through fan cooling with no pointless PCB behind it...think again.

July 15, 2015 | 08:05 AM - Posted by obababoy

I think it is a great idea. straight through fan cooling with no pointless PCB behind it...think again.

July 10, 2015 | 03:56 PM - Posted by StephanS

AMD priced this wrong.. unless they have very little volume.

At $499 , this card would completely overtake the GTX 980.
It would be even hard for nvidia zealot to get a GTX 980 knowing they paid the same amount of $ for a slower card.

But at $550, AMD just left a huge market for nvidia.

I can only guess this : AMD doesn't want to anger nvidia.

Because nvidia got like 8 billion in cash in the bank and they can price their card so low that AMD would be loosing money if they tried to compete...

The Fury is a nice card (super quiet, very low temperature, faster then a GTX 980), but a still 50$ overpriced.

Now, if the Fury card was $550 and including all 4096 shaders,
AMD would totally OWN the gaming GPU market... (untill nvidia drop the 980 ti to $500)

AMD, destined to eat the scraps of Intel and Nvidia ...

July 10, 2015 | 05:41 PM - Posted by arbiter

Well only reason its priced as it is was cause 980ti and its price point. AMD was forced to price this lower then they wanted to.

Look at benchmarks between 980ti and fury you can see why they had no choice.

July 10, 2015 | 05:46 PM - Posted by Anonymous (not verified)

The 290x has recently been restested and is not beating the 980 in most games. Keep an open mind as AMD cards age well as the driver matures. My 7970ghz Crossfire is beating 980 easy.

July 10, 2015 | 07:15 PM - Posted by Marcelo Viana (not verified)

Why no reviews take on that IQ issue on nvidias card?
just give fps isn't enough.
Must put everithing on maximum IQ and found a way to get the same IQ on nvidias card, because is already know that nvidia cripple their cards IQ to get more FPS.
So for us that intend to game 4K and or buy a expensive card must put IQ over FPS.
If the benchmarck is all about FPS then give the video of each test on the review pages for us to see the diference in order to choose fairly.

July 11, 2015 | 02:13 AM - Posted by capawesome9870

@Ryan-Shrout

Could you water cool the Fury, from the ~70c to the 50c of the FuryX?

then Check Wattage of the Fury?

from what we heard before the fury launched it was adding a water cooler to help keep temps low to also help reduce the wattage leakage.

July 11, 2015 | 09:37 AM - Posted by Topjet (John) (not verified)

Very good job on the review, liked it :) What it tells me is 4k gaming is going to take a lot of money, SLI or Crossfire, to me it looks like AMD should let it's vendors/partners make a air cooled Fury X at $599.99 lol

July 12, 2015 | 04:28 AM - Posted by HERETIC (not verified)

Hi Ryan.
With everyone wanting"MOAR"and AMD pushing close to the limit
of this architecture to compete, have you considered trying some under-volting?
Seems the Asus card on a few sites drawing 30 to 50 Watts
less than the Sapphire...................

July 12, 2015 | 03:35 PM - Posted by StephanS

I agree. My big thing in the past 5+ year has been undervolting and finding a chi sweet spot.

The 290x is a great example, as it shatter all the TDP result we see on reviews.. if you dont mind running it 5% to 10% lower clock.

AMD got great chips, but doesn't seem to read tech site review to at least make one product that wont get destroyed by the press.

"Its to hot", "use to much power", "Not recommended"
Then a few month later an nvidia card comes out that use more power, it hotter, but its receives awards after awards...

No one seem to be looking at price/performance, where AMD 'used' to be a clear leader.
With this Fury release, AMD now eliminated their only advantage... Who in the world in running this company to be so out of touch with simple concepts.

July 15, 2015 | 08:08 AM - Posted by obababoy

I totally agree. I feel like NO ONE at AMD really takes advice from normal everyday gamer's when it comes to GPU's. AMD's only advantage was as you said price/performance, which they still have on the 290/290x. These new cards are great performers but the price just isn't going to work for AMD.

July 13, 2015 | 01:35 PM - Posted by Anonymous (not verified)

I'm really surprised you didn't use Witcher 3 for these tests. That game would be one of the best to test the mettle of these latest cards.

July 15, 2015 | 10:38 AM - Posted by gerard (not verified)

I think the obvious question is as the sapphire version uses what appears to be a fury x pcb is it possible to unlock the core with a bios flash? More than likely its a fury x with some shaders disabled via a bios, seeing as the card has a dual bios it shouldn't be that hard to test?

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.