• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

first AMD Radeon Pro Duo ReviewWorld’s Fastest Graphics

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
They are comparing it to the fastest cards currently available.

If all of a sudden you are expecting the dual GPU beast to be better value than the 2x single cards are, you are in the wrong era. There's a premium for the top single card. It's been like that for a while now.

I never thought that dual GPU cards had good value to begin with. To me, this Radeon Pro Duo is fundamentally the exact same thing as both the R9 295X2 and the Titan Z: a bunch of overpriced halo cards.
 
yeah i'm in the process of putting together a mITX build... so this card seems pretty attractive to me... but with Pascal right around the corner... I may have to wait. I have to drive 3440x1440p at decent framerates now... I've been holding off on the 980Ti for a bit, because it can't drive that resolution decently... and the Titan X isn't any better...
 
yeah i'm in the process of putting together a mITX build... so this card seems pretty attractive to me... but with Pascal right around the corner... I may have to wait. I have to drive 3440x1440p at decent framerates now... I've been holding off on the 980Ti for a bit, because it can't drive that resolution decently... and the Titan X isn't any better...

You have to wait for big Pascal or Vega 10 for that kind of grunt in one-slot.

When they arrive in 2017, they could be very expensive!
 
And that is precisely why this card is DOA. The marketing spin of "dude, so good for VR" is just the nonsense they are using to try to sell this hugely overpriced card built with year-old GPUs at inflated prices relative to two Furies in CF.

Had this come out around when Fury/Fury X launched I could sorta kinda see a reason for its existence. Today, it is truly a product that's just being put out because AMD doesn't want to deal with the negative optics of cancelling a product that they pumped at E3 2015 as "World's Fastest Card."

Of course you are entitled to your opinion as to why they released it during the release of VR. I tend to look at future reasons. Planting the seed for multi gpu not only for VR but for devs for DX12 and Vulcan. There is this little NAVI thing that will hit the market in 2018. NAVI by definition is a Variable star both intrinsic and extrinsic. Dual Vega or Polaris or APU + gpu on interposer with HBM2 or next gen memory will be what this is about IMO. Devs now have a card that they can use to design their next game engines on. Even Nvidia will be looking towards this type of platform because of NVLink. You may not realize or want to realize that AMD is the company moving the gaming industry forward while Nvidia is moving towards a major opportunity with AI and computerized cars.
 
You have to wait for big Pascal or Vega 10 for that kind of grunt in one-slot.

When they arrive in 2017, they could be very expensive!

That's why this card seems that it'll do me pretty decently... With Tomb Raider almost hitting 60 FPS at 4K, and with 3440x1440p at half the amount of pixels, it should be decent performance... (Maybe... depending on drivers)
 
It's not tripe, this Pro Duo actually can use Firepro drivers.

16Tflops at this power level blows even GP100 away for raw performance and perf/w. If you don't need CUDA and need a HPC capable accelerator, this thing makes a lot of sense.

For gaming? Not so much sense outside of mITX and slot-limited scenarios.

16TFLOPs at single precision is impressive. But its DP rate rather low not to mention being only 8GB in total (4GB per GPU) if your talking HPC. Not really comparable to a P100 because they are for different markets.
 
16TFLOPs at single precision is impressive. But its DP rate rather low not to mention being only 8GB in total (4GB per GPU) if your talking HPC. Not really comparable to a P100 because they are for different markets.
It can be used for 3d modeling, cad/cam softwares, opencl acceleration (eg video transcoding or rendering) etc
 
I never thought that dual GPU cards had good value to begin with. To me, this Radeon Pro Duo is fundamentally the exact same thing as both the R9 295X2 and the Titan Z: a bunch of overpriced halo cards.

R9 295 X2 was about $500 in mid-2015. At that price, it was a pretty good buy, considering the raw power and excellent build quality.
 
It's not tripe, this Pro Duo actually can use Firepro drivers.

16Tflops at this power level blows even GP100 away for raw performance and perf/w. If you don't need CUDA and need a HPC capable accelerator, this thing makes a lot of sense.

For gaming? Not so much sense outside of mITX and slot-limited scenarios.

I'm also thinking of a graphics workstation card. The W8100 @ $999 is a bargain. This is ~2x as good.
 
Of course you are entitled to your opinion as to why they released it during the release of VR. I tend to look at future reasons. Planting the seed for multi gpu not only for VR but for devs for DX12 and Vulcan. There is this little NAVI thing that will hit the market in 2018. NAVI by definition is a Variable star both intrinsic and extrinsic. Dual Vega or Polaris or APU + gpu on interposer with HBM2 or next gen memory will be what this is about IMO. Devs now have a card that they can use to design their next game engines on. Even Nvidia will be looking towards this type of platform because of NVLink. You may not realize or want to realize that AMD is the company moving the gaming industry forward while Nvidia is moving towards a major opportunity with AI and computerized cars.

Getting your hardware into a devs workstation is huge. nVidia's been doing it for years. They basically sent complete VR workstations to leading devs. Real smart move considering they are already working with their hardware in the consoles.
 
SOURCE Per TechPowerUp
91a.jpg
 
I guess the reviews are just so bad that they want people to make mindless decisions on purchasing this 1500.00 GPU hoping that the reviews stop them..
 
Isn't that a contradiction ?

Fury X is already clocked to the limits and it won't beat a 980ti / titan x overclocked at 4K..

At high resolution I see high fps increase from the memory overclock alone.

So tell me why Fury X actually is actually fastest reference GPU in the world according to Techpowerup suite?

If we compare reference models of GPUs Fury X in 4K is faster than any other single GPU out there.

Go to recent review of any GPU on TPU and count the games. Fury X is faster in 4K than Titan X in 10 out of 15 games in their review suite. 10!
 
Back
Top