Slashdot Mirror
Air-Cooled AMD Radeon R9 Fury Arrives For $100 Less With Fury X-Like Performance
MojoKid writes: When AMD launched the liquid-cooled Radeon Fury X, it was obvious the was company willing to commit to new architecture and bleeding edge technologies (Fiji and High-Bandwidth Memory, respectively). However, it fell shy of the mark that enthusiasts hoped it would achieve, unable to quite deliver a definitive victory against NVIDIA's GeForce GTX 980 Ti. However, AMD just launched their Radeon R9 Fury (no "X" and sometimes referred to as "Fury Air"), a graphics card that brings a more compelling value proposition to the table. It's the Fury release that should give AMD a competitive edge against NVIDIA in the $500+ graphics card bracket. AMD's Radeon R9 Fury's basic specs are mostly identical to the liquid-cooled flagship Fury X, except for two important distinctions. There's a 50MHz reduction in GPU clock speed to 1000MHz, and 512 fewer stream processors for a total of 3584, versus what Fury X has on board. Here's the interesting news which the benchmark results demonstrate: In price the Fury veers closer to the NVIDIA GeForce GTX 980, but in performance it sneaks in awfully close to the GTX 980 Ti.
Depends on the resolution. 1440p and even 4k monitors are becoming more common and you need the extra power. It will be interesting to see how Dx12 will impact performance too.
Wait for Ubuntu 16.04 (then three monthes after, you'll long for Ubuntu 18.04 but sssh...)
Don't forget high-fps monitors, which is an even harder problem to solve than resolution. I've got an overclocked Geforce 980 Ti (faster than the Fury X) and it still can't manage 1080p @ 144hz in many games.
Apple has thousands of developers, artists and other experts that get paid to work on OSX.
When do you think a Linux desktop company will be come close to matching that?
Granted, most of the work that they do at Apple is thrown away before it reaches the consumer, but that is often the nature of product development, when you don't know which features the users will need or want. The same would be true for a Linux desktop OS company.
..and just like Nvidia, still using .28nm process for the GPU, same as it has been since ~2010-2011... by "technology standards", this is and incredibly long period of time. I totally understand the issues with supply from TSMC, Samsung, etc, and that the products of "latest-and-greatest" in chip fabrication are supplying the smart phone and tablet industry as fast as they can.... But my point is that these days, every time I see either AMD or Nvidia releasing yet another hot and power-hungry rehash, I sadly shake my head. Gigs and gigs of RAM are great - so is 1440p and 60 fps, but I want lower-power consumption and I want less heat. I don't want increasingly complicated cooling solutions.
Perhaps for current-day games, but the proposed specifications for the commercial Oculus Rift are quite high (and that's just the "recommended" specs): https://www.oculus.com/en-us/b...
The high-end cards of today will be the mid-high range cards of next year, so I wouldn't be surprised if some of the more demanding VR games make full use of the available power.
Does it have good Linux drivers? I.e. that have the same performance, memory requirements etc. that the windows counterpart? Doesn't have to be free, only good.
No? Then I'm not interested...
Stefan Axelsson
Benchmarks mean nothing when so many games have extra bugs with AMD products. They really need to do something about building up a reputation for stable drivers that offer stable performance, even for new games. Because right now, raw performance means dick to the customers when it comes to crashes and poor FPS.
As soon as you write a driver for it.
It's open source, after all. Nothing stopping you.
If you're so loaded that you're buying a 50 inch 4k TV and a $500+ graphics card, I'm sure you can shell out $20 for an adapter. http://www.amazon.com/dp/B00E9...
That's because manufacturers run into limits, especially around cost, since Moore's law has reached the end of the line. A transistor on 20nm or 14nm is more expensive than a transistor on 28nm.
thegodmovie.com - watch it
I'm starting to think I'm getting old and am the only person who doesn't give a shit about the Occulus Rift or any other VR setups, at least not yet. Seriously, I've always been a graphics junkie, since the days when CGA was the standard for color, and I really am quite happy right now with 1080p on a flat screen.
I've got a 1440 and can quite easily have a shitty experience if I select the wrong settings (2x770GTX 4Gb).
Still, I'm happy as larry that the PC world has finally decided to leave 1080 panels behind. I was running higher res than 1080 for years, and then those pesky TV panels turned up everywhere and put us back years.
My old-ass eyes can barely tell the difference between 1080 and 4k. Give me a nice big monitor, and a game that runs smoothly (which apparently is hard for some companies *arkham knight*) and I don't really need to spend the money on two Titans. Who decided that we need photorealism in games, anyway?
You are welcome on my lawn.
Yeah, if you don't like it, talk to TSMC (and maybe Samsung), and the people that supply them. This isn't NVIDIA's fault, and AMD/ATI is in the exact same boat, so quit whining.
I'd like to know where you're buying a 50" 4k TV for 600$...
I've got better things to do tonight than die.
Fancy ass Arkham City didn't need a $200+ video card.
You are welcome on my lawn.
Correct. There is no such thing as an adapter. At some point someone may make a display port to hdmi one, but at the moment there is no way to get 4k at 60hz on an hdmi tv except to buy an Nvidia card. I researched a ton and then ended up selling my existing AMD card and buying a 970 when I got tired of 30hz.
On that note my 4k TV is a 40inch and cost ~600. I wouldnt reccomend more than 40 for a desktop monitor. Even at 40 you have to turn your head a lot.
Not to mention the cores are huge so yield rate is likely very touchy.
Who decided that we need photorealism in games, anyway?
The developers did, to make up for not really adding any new gameplay or content. They want to sell 'the next big thing' without really needing to do anything but reiterate the old stuff at higher rez.
What is really interesting to me about these aircooled Fury cards is that even though the PCB of the card is much shorter than that of a typical flagship GPU card the heatsinks being use extend the card length out to the typical 12" length. Why is that interesting you say? the power consumption of the card is on par with other AMD GCN cards and when it comes to dissipating the associated heat it still requires the same mass of copper and aluminum fins to avoid temperature spikes the associated fan acceleration. On a long enough timeline it seems to me the reliability of this card will be noteworthy in comparison to other cards using a more typical memory and cooling configuration.
The Fury is really only competitive at 4K resolution. At lower resolutions 1440p, 1080p, etc., it gets beat pretty bad in pretty much every game out there (save for a very small handful) by the 980 and 980 TI. Given that the majority of monitors out there are still 1080p or 1440p it is hard to recommend this card.
We were all warned a long time ago that MS products sucked, remember the Magic 8 Ball said, "Outlook not so good"
Comment is too short.
I guess it's just the way the engineering played out. Basically everyone except Intel is stuck on 28nm, so there's not much to work with... might as well pay your engineers to rig up a way to deal with all that heat. At least AMD is putting effective coolers on their cards instead of nVidia just putting the "prettiest" blower on there.
I don't really share your want for lower-power graphics cards though. These are *desktop* parts connected to the electrical mains. I don't live in communist germany where electricity costs 3-4x as much as it should, the cost is negligible for me; all I care about is maximum performance per unit cost; if that means a 1.2 kW system then so be it!
Aye, more like a $25 or $50 card... or hell,$0 with on-board graphics on a laptop.
It's not a notebook GPU. It is a desktop GPU. Why would you be worrying about power consumption and heat? This is marketed toward PCs.
Uhhhh you DO know there are 4K monitors with displayport, yes? If you are plugging in a PC I'd suggest sticking with monitors, I've set up plenty of HTPCs for customers and frankly I've yet to see a TV that a monitor at the same res didn't look and/or perform better.
ACs don't waste your time replying, your posts are never seen by me.
Oh, you're paying for the on-board graphics. You can't NOT pay for it, because Intel welded it to the CPU.
Sleep your way to a whiter smile...date a dentist!
So turn off smoothing. It's a function of the tv, and takes about 10 seconds to turn off. I personally hate the feature, it makes movies look like they were filmed for basic cable tv.
"Well kids, you tried your best, and you failed. The lesson is, never try." -Homer Simpson
You should get some cough mixture for that cough. It sounds terrible.
I have determined that my sig is indeterminate.
In the last years, Nvidia have made big strides in reducing their power consumption for a given performance. You can buy the "latest-and-greatest" in performance, which will outperform older cards, OR you can get similar performance in a smaller, cooler and cheaper package. The 750Ti comes to mind:
It is "only" a midrange card, but with a power consumption of 60-70W it does not even need an additional PCIe power connector.
Recently, AMD are also getting closer with HBM on the Fury (although they are still falling a bit short of Nvidia).
If you think back a few years, the roles were reversed BTW:
Nvidia was still on the Fermi (also derided as "Thermi") architecture and significantly less efficient than AMD's HD5xxx series.
Looking forward, the .14nm process is supposed to come out in 2016 and HBM is supposed to get its 2nd generation. I think those will be good times.
C - the footgun of programming languages
as most 50 inch+ 4k tv's don't come with with displayport, but with a hdmi 2.0 port, and 4k pc monitors don't come in usefull sizes and prices. This card isn't worth looking at for me.
DisplayPort contains HDMI, this is like complaining a your laptop doesn't have a microUSB port and therefore you cannot connect it to a phone. Just get a dumb DP to HDMI cable.
It's not a notebook GPU. It is a desktop GPU. Why would you be worrying about power consumption and heat? This is marketed toward PCs.
BECAUSE I DON'T WANT TO HAVE TO SHOUT OVER ALL THIS FAN NOISE!
Well, the noise issue is mostly solved with aftermarket coolers, but that still leaves power consumption and heat. I guess none of this matters for the occasional gamer, but if you do productive work on GPUs 24/7, and (gasp) pay for your electricity, then these things matter.
(I've been building silent, often fanless computers since about 2003, since I simply don't want any extra noise where I live. Besides, I've never understood why it's OK to waste energy willy-nilly just because it's plugged in. Most of my computers have "laptop" parts in "desktop" cases for the best combination of low power consumption and cooling.)
Escher was the first MC and Giger invented the HR department.
"BECAUSE I DON'T WANT TO HAVE TO SHOUT OVER ALL THIS FAN NOISE! "
Son, unless you're running Delta fans, you have no right to say shit about noise.
I've got a single delta fan louder than a QUAD SLI TITAN setup.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
" i doubt a regular vga cable can transport 4k at 60 hz.. because i know it allready has problems with full hd.."
Bullshit. I was doing OVER Full HD (That would be 2048x1536) on VGA OVER A DECADE AGO on a 21" Trinitron.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Next year is the year of the Linux Desktop.
x
I usually go by the performance index of www.3dcenter.org, which gives an average performance value relative to the Radeon HD 5750/6750 GDDR5, which is defined as 100%.
The index is not based on theoretical GFLOPS, but on tests by various review sites (mostly gaming) and calculated for benchmark results at 1920x1080 with 4x multisampling anti-aliasing.
This explains why Nvidia looks better in the 3dcenter.org ranking, as they usually get more gaming performance out of cards with the same GFLOPS.
3dcenter.org also calculates a performance/watt rating where they divide the performance index by the typical power consumption in games. The result is in percent of performance per watt, and as explained above it favors Nvidia. Of course, if you do something other than gaming, your results may differ.
The best result at the moment is for the GTX 980 4GB at 3.45, closely followed by the 750Ti at 3.44. I used the 750Ti as example of a midrange card that still performs quite nicely compared to high end cards of a few years ago. Current market price is 130-145 Euro. The Fury X is listed with a performance per watt of 2.32.
BTW, Wikipedia says that
Full-height cards may increase their power after configuration. They can use up to 75 W (3.3 V Ã-- 3 A + 12 V Ã-- 5.5 A)
Graphics cards manufacturers use that routinely to save a few cent on the extra connector.
C - the footgun of programming languages