Comments Locked

63 Comments

Back to Article

  • teldar - Tuesday, April 26, 2016 - link

    I'm not a huge fan of the card, but you guys act like this is something AMD is targeting the masses with. I think the very name of the card indicates that it's meant for VR content creation rather than comsumption, but you still talk about it as if it is something being pushed for end users.
  • tuxfool - Tuesday, April 26, 2016 - link

    The Titan cards are in a similar position, yet consumers still buy that card, value for money or no.
  • Strunf - Tuesday, April 26, 2016 - link

    Virtually no one is buying the Titan... its not even on Steam stats and the GTX 980 has close to 1% market share.
  • LostWander - Tuesday, April 26, 2016 - link

    "A lot" of sales for a card as expensive as the Titan wouldn't really be enough to show up on those stats. If I'm reading it right the entire R9-200 series is only a bit above 1% itself. Its a small group of people compared to the rest of the PC field that buy cards in this price range but they aren't insignificant.
  • Murloc - Tuesday, April 26, 2016 - link

    on steam most players have integrated intel GPUs so I don't think those statistics are relevant to the crowd that is in the intersection between Anandtech and gaming.
  • jabber - Wednesday, April 27, 2016 - link

    Yeah cards like the Titan sell in the hundreds to low thousands. Its like the S class Merc. What you see in the S class is what will be in standard cars in the near future.
  • P39Airacobra - Saturday, May 14, 2016 - link

    Actually the majority of PC gamers do not use steam. They still buy them in physical copy. If you look you will see a small percentage of steam gamers use a gaming GPU or APU, It's mostly IGP. Most steam users play indie games that can run on a old core duo. In other words the majority of steam users are not really considered gamers. But I agree that the majority of gamers don't buy overpriced high end. And $1500 dollars is insanity for a GPU that will be priced at $500 or $600 after it sits on shelves because everyone is buying a 1070 or 1080 instead.
  • Husky42 - Monday, May 30, 2016 - link

    This has to be one of the dumbest comments I have read.. Most gamers do not use steam yet most games use steam... So I guess you do not play games. Gaming for 24 years and have had steam since orange box. I'm actually wondering what you think buying a physical copy does for you.. it does nothing considering you still cannot install the phsyical copy of a game without steam.. Steam is something the MAJORITY of gamers are FORCED to use, but they still use it.

    746,575 962,426 Dota 2
    554,590 584,387 Counter-Strike: Global Offensive
    87,187 98,969 Total War™: WARHAMMER®
    67,608 70,734 Team Fortress 2
    66,087 69,843 ARK: Survival Evolved
    58,341 61,596 Football Manager 2016
    56,694 64,351 Sid Meier's Civilization V
    48,672 51,320 Garry's Mod
    46,578 51,892 Fallout 4
    37,553 42,403 Warframe

    Um.. not a single indie game in the top played for today.
  • Husky42 - Monday, May 30, 2016 - link

    http://store.steampowered.com/hwsurvey/processormf...
  • Husky42 - Monday, May 30, 2016 - link

    Gonna continue talking our your derriere?

    http://store.steampowered.com/hwsurvey/videocard/
  • AS118 - Tuesday, April 26, 2016 - link

    To be fair, I feel like more people buy the 980Ti rather than the Titan. With OC-ing, the 980Ti can catch up to the Titan anyhow, so personally, I would rather get a 980Ti or Fury X rather than a Titan X.
  • lilkwarrior - Thursday, April 28, 2016 - link

    The Titan X is a *prosumer* card; this card is also pretty much that as well. The Titan X has precision/compute capabilities (& memory capacities) completely irrelevant to gamers.

    It's meant to be for pros who want to their home computer away from their workstation at work be somewhat decent job doing that while still being able to play games well.
  • Denithor - Thursday, April 28, 2016 - link

    Actually it isn't. It was marketed as such, but it has the same 1/32 FP32 as the 980/Ti offered so it's no better than gamer level cards for high level crunching. Contrast this to the 1/3 FP32 offered by the previous generations (Titan & Titan Black) and you'll see the difference.

    http://www.anandtech.com/show/9059/the-nvidia-gefo...
  • milkod2001 - Tuesday, April 26, 2016 - link

    This card is for 4k gaming / media content creation / VR etc. It's not only for VR content creation.

    This is card to get if you are not willing to wait 8-9 months for new 980t i/ titan replacements i guess
  • milkod2001 - Tuesday, April 26, 2016 - link

    t
  • superbn3rd - Wednesday, April 27, 2016 - link

    this is the card you'd be wasting money on.... both pascal and vega 10(amd claims there new cards will be over 2x the performance of there previous cards) will be out later this year with 16gb of hbm2 vram... honestly not a great time to upgrade pc.. to many releases around the corner(xpoint, zen, vega 10, pascal).
  • Alexvrb - Friday, April 29, 2016 - link

    They probably won't be HBM2, actually, from what I'm reading that's coming in 2017.
  • CiccioB - Tuesday, April 26, 2016 - link

    This is the card to get if you have lots of money to throw away and you want lots of unsupported games in next few months. Crossfire/SLI solutions are always the worst solution for trying to get better frame rates.

    Moreover I wonder how you can develop with that small VRAM memory quantity. You cannot fill the entire resource set in that RAM if you have also the development and debugger tools active.
    I wonder how that can be targeted at VR content creation.
  • defaultluser - Tuesday, April 26, 2016 - link

    Yeah, usually "development" boards have twice the ram. But with the chip saddled with HBM1, that's not an option here.

    So the plan is to market a pointless overpriced card at whoever will be stupid enough to bite?
  • vladx - Tuesday, April 26, 2016 - link

    You do know VR is not just about gaming, think Hololens and other applications that could be used in tourism, entertainment and even automotives. Unlike gaming, those don't need huge amounts of VRAM to develop.
  • AS118 - Tuesday, April 26, 2016 - link

    That's a good point, Crossfire / SLI has always had hit and miss support, but lately I feel like it's worse than it's been in a while, and many major releases don't support it or don't support it well enough for it to be worth it.

    Especially with all the bad console ports that bring top-tier cards to their knees because the developers just couldn't be bothered to port it right, even when the original game was developed for a x86 AMD APU platform that's basically just a modified PC based on older tech.

    I'll admit that even with a x86 common platform it's not exactly the same, but even still, some companies seem to not even want to try, including Microsoft (the makers of Windows) of all people!
  • beck2050 - Wednesday, April 27, 2016 - link

    Water cooling AND Crossfire? Pass. 2 custom 980tis are a much better and cheaper solution if going dual, which is never that great to begin with. This is just a marketing gimmick.
  • HammerStrike - Thursday, April 28, 2016 - link

    "...alongside the FirePro drivers for content creators, Radeon drivers will be available. The Pro Duo will not see validation for as many applications as a true FirePro card, but official support will be provided for applications important to gaming content creation such as Autodesk, Maya, and Blackmagic Davinci Resolve."

    The (limited) FirePro driver support is the only reason to consider this card, unless you have absolute space constraints in a SFF chassis and want the option of crossfire.

    That being said, $1500 for a professional (or prosumer, in this case) card is on the cheap side, as these things go. Also, would be interesting to see if driver support for the listed programs includes crossfire support for those applications - you'd have to think it would give the nature of the card. If that's the case that's actually pretty interesting, as there is no SLI/Crossfire support in the professional space. If AMD is supporting multiple GPU's for those applications there could be some interesting niche cases where the raw performance of dual Fury X GPU's could be of use, memory limitations notwithstanding.
  • GerardFreeman - Tuesday, April 26, 2016 - link

    You mean the 1080Ti dont you?
  • MrSpadge - Tuesday, April 26, 2016 - link

    Are they? I'm reading in the 2nd/3rd sentence:
    "... releasing the Radeon Pro Duo to the market as their first card targeted at VR developers. This card is not being directly aimed at gamers"
    And in the next paragraph:
    "... the card is being marketed for VR content creation first and foremost"
    I have no idea how you can read this as "AMD is targeting the masses".
  • Cellar Door - Wednesday, April 27, 2016 - link

    No one cares if you are a fan or not. There is no actual point in your response that has not been addressed in the article.
  • MUATAZ - Tuesday, April 26, 2016 - link

    when we expect to see High end product with 14nm ?
  • T1beriu - Tuesday, April 26, 2016 - link

    Vega 10 - Very late 2016/early 2017.
  • nevcairiel - Tuesday, April 26, 2016 - link

    Do people really expect AMD to release Polaris now and Vega in the same year?
  • MrSpadge - Tuesday, April 26, 2016 - link

    Vega is probably the same architecture in a bigger chip with HBM2, so there's no technical hurdle to release as fast as possible. You can argue that in this case they won't have enough new cards for 2017, but being quick has always payed off in the GPU business. (I'd rather bet on 2017 due to the card not being ready enough in 2016)
  • CajunArson - Tuesday, April 26, 2016 - link

    So apparently AMD is being very stingy with review samples and even the TechReport, whose founder is now an AMD employee, didn't get a review sample.

    Can we take it from this story + the lack of a review that Anandtech didn't get one either?
  • silverblue - Tuesday, April 26, 2016 - link

    I haven't seen any reviews in the wild as of yet.
  • Ryan Smith - Tuesday, April 26, 2016 - link

    We're not getting a review sample either. And I'm fine with that. As a gaming card I'd argue your better off with a single GPU setup at this time, and for content creation we aren't currently setup to review VR content creation and the other kinds of use cases AMD is pitching for the card.
  • Foxorroxors - Tuesday, April 26, 2016 - link

    you gotta admit, that first picture is geek porn.....
  • TEAMSWITCHER - Tuesday, April 26, 2016 - link

    These ridiculous Radiator designs are laughable..
  • Creig - Tuesday, April 26, 2016 - link

    ...until Nvidia comes out with one of their own. Then, suddenly, it will be the "right time" for AIO liquid cooled GPU designs.
  • xthetenth - Tuesday, April 26, 2016 - link

    If by laughable you mean by far the best reference coolers out there and/or make a sick joke out of the sad sack blower(!) NV's been saddling their top chips with, then yes. Otherwise, what do you have against good cooling solutions?
  • MrSpadge - Tuesday, April 26, 2016 - link

    "Even with the price, AMD is keen to admit that the Radeon Pro Duo is now the single fastest graphics card on the market"
    It must have taken them a lot of willpower to admit that ;)
  • BurntMyBacon - Wednesday, April 27, 2016 - link

    @MrSpadge: "It must have taken them a lot of willpower to admit that ;)"

    Maybe they were afraid that nVidia had a card sitting on the bench, just waiting for them to claim the top spot so that they could rain on their parade ... Yeah, I got nothin.
  • TesseractOrion - Tuesday, April 26, 2016 - link

    Do you think the price will fall dramatically quite quickly as it did with the 295X2? Seeing how new cards are coming out so soon...

    Was hoping for a price of less than 2 Furies otherwise not sure of the point (bar convenience of one big card instead of two of course)....?
  • haukionkannel - Tuesday, April 26, 2016 - link

    Most propably near the launch of Vega... This is expensive card to produce, so early 2017 there may be price cut. Most propably not.
  • blzd - Tuesday, April 26, 2016 - link

    Dual GPU cards have always come with a price premium. Just not usually at this level.
  • Death666Angel - Tuesday, April 26, 2016 - link

    Their prices fell, really? The times I checked them (for potentially driving a VR HMD), they were all above 900€, about 3 times the cost of a R9 290. Sure, they probably dropped in price compared to their MSRP, but then again, everything does.
  • webdoctors - Tuesday, April 26, 2016 - link

    Man, what monkeys are running this division? Are they actually going to recoup the cost of this board? The volume already for these cards is so low, when you have to engineer this SLI frankenstein and than do the verif and QA checks. How likely are you going to recoup costs when there's going to be a GPU refresh in months so the window for this card to sell is tiny.
  • Death666Angel - Tuesday, April 26, 2016 - link

    "How likely are you going to recoup costs when there's going to be a GPU refresh in months so the window for this card to sell is tiny."
    Dual GPU cards have usually ended up being release 6 months to a year after the initial release. So even if we get new GPUs in the coming months, a new dual GPU card won't happen for another good chunk of time.
  • Beany2013 - Tuesday, April 26, 2016 - link

    In the professional world, time is money.

    In some markets, the time is a *lot* of money.

    The cost of these cards could be saved just by knocking a few hours off a finished project in some areas, if it means you can move onto the next project sooner - or get the deposit for the next project sooner, meaning you can expand your business with the extra cashflow, etc.

    Whether it'll be replaced in six months time with something just as quick using less power is irrelevant; whether it can give me a competitive edge *now* so I can take that upcoming contract from the competition is what matters in most of the professional hardware purchasing world.
  • webdoctors - Tuesday, April 26, 2016 - link

    Beany2013, I agree with you completely. Folks will buy it, but will it be enough to break even?
    Assuming there's 100 VR companies (I'm assuming that's a lot of VR companies?), and each company buys 20 cards, that's 2000 cards. That's $3M. If these have a 70% profit margin, than they make $2M. That'll pay for 10-15 engineers depending on where they're based...I'm guessing this chip took 6 months time for 30-50 engineers. Or cost of 15-25 engineers...

    So on the surface it seems this is a loss leader? Maybe to get devs hooked on using AMD products for VR development?
  • prtskg - Wednesday, April 27, 2016 - link

    You have forgotten that there is a firepro version of this card too which has already won a contract in chime telescope. I wouldn't be surprised if the investment in dual fury cards would be recouped faster than other fury cards.
  • webdoctors - Friday, May 6, 2016 - link

    With the release of the 1080 card at the end of the month for half the price or less of this card, this card will be obsolete before launch.
  • onbquo - Tuesday, April 26, 2016 - link

    oh boy! Have you seen that 1 TB/s memory bandwidth? I think that's a first..
  • abufrejoval - Tuesday, April 26, 2016 - link

    So why would I not simply use two Nanos? Because there is only one HDMI connector on the VR headset? So the 500 buck premium is about multiplexing a signal which will get de-multiplexed on the headset? Seems a bad deal, in a way, although I'd rather like to have 0 cables instead of 2 and I guess that would be worth $500 extra....

    I had a Nano for about two weeks until I got cold feed about slow frame rates on Prepar3d using the Oculus plug-in (turned out it wasn't a Nano issue) and returned it while I could for a full refund.

    I must say at €500 it was quite (and quiet) a nice card and exactly the value it failed to be at €700.
    Yes the GTX 980ti which replaced it is as much faster as it is more expensive, but physically a brute compared to that Nano cutie...

    A dual Nano sounds good, but I'd only get interested once it drops to 2x Nano price.

    Seems Avago hiked those PLX prices after they bougth the company to pay for the purchase and that is perhaps a large part of the premium.
  • vladx - Tuesday, April 26, 2016 - link

    "So why would I not simply use two Nanos? "

    One word: latency. That's more important in VR development than price.
  • samer1970 - Tuesday, April 26, 2016 - link

    the same latency , this card uses a plx switch chip which splits the 16x pci express into 2 virtual 16X slots ... this is actually SLOWER than using two REAL 16X slots on an X99 motherboard !

    two Nanos in CF is FASTER than this card .
  • Beany2013 - Tuesday, April 26, 2016 - link

    It might be marginally slower, but you need to consider heat, power, compute density and complexity - of which this card should be far less of a concern on each point than a crossfire setup.

    Those are things you avoid in the professional world where stability and uptime are king.
  • samer1970 - Tuesday, April 26, 2016 - link

    this IS a cross fire setup. inside the card ... it is simply the 16X slot divided into 2x16 slots using plx switch chip , namely the plx 8747 , which connects to the cpu telling it that it has 2x16 slots instead of one and gives electrical 2x16 slots connection for 2 GPU on that board , the design on the board is a CF design 100% each chip is connected to the PLX chip via 16 virtual lanes (look at it as if there are 2 virtual lots now used ...

    some motherboards include this plx chip to give you virtual 2x16 slots ... it is the same 100%
  • svan1971 - Tuesday, April 26, 2016 - link

    I'm sorry but isn't a new chip due out this summer that's promising a 50% speed increase doubling the current ram, running cooler and more power efficient? If your buying this card you have money to burn.
  • Gigaplex - Tuesday, April 26, 2016 - link

    "At a high level, the Pro Duo should give us up to twice the performance at twice the power consumption (plus a bit extra for PCIe switches)."

    The rated power consumption has "only" jumped from 275W to 350W. Hardly twice the power consumption.
  • Ryan Smith - Tuesday, April 26, 2016 - link

    R9 Nano is 175W, not 275W.
  • beck2050 - Wednesday, April 27, 2016 - link

    Not available to independent review sites. Yeah that's encouraging.
  • BurntMyBacon - Wednesday, April 27, 2016 - link

    @beck2050: "Not available to independent review sites. Yeah that's encouraging."

    Which independent review sites are setup to review VR development? I'm giving them slack on this one.
  • BrokenCrayons - Wednesday, April 27, 2016 - link

    I hate to say it, but cramming two graphics cards on a single PCB is a really inelegant solution to the problem that VR creates for content developers - a problem that I just don't see being a persistent industry issue because of the small probability that VR will become an enduring, widely adopted technology.
  • K4AGO - Sunday, May 1, 2016 - link

    You didn't hate to say it as much as I hated to read it. Intelegant... Very intelegant !
  • ZycaManiac - Sunday, May 1, 2016 - link

    This is just such a weird product, released at a weird time...

    I'm quite sure whoever buys this will likely be pretty pissed about the new products that they will see in 3 month...

    Then again, if have enough disposable income, they might just buy the new card...

Log in

Don't have an account? Sign up now