D
Point of the story.
The point of the article is about more than overtessellation in one title. The point of the article is that closed libraries have the potential to create even more of a walled garden GPU effect.
Right now, a conventional "Gaming Evolved" or "TWIMTBP" title ships out optimized for one vendor but can still be optimized for the other post-launch. GameWorks changes that.
I consider this problematic because we've seen how companies can abuse this kind of power. 12 years ago, Intel began shipping versions of its compiler that refused to optimize for AMD hardware, even though AMD had paid Intel for the right to implement certain SIMD instruction sets. This fact wasn't widely known for years. Instead, people concluded that AMD's implementation of the various SIMD sets must have been sub-optimal, because it didn't benefit from using SSE or SSE2 the way Intel did. Since K8's implementation of SSE2 was only 64-bits wide, the conclusion was that AMD had fumbled the ball in that regard. In reality, Intel's compilers would refuse to create the most advantageous code paths for AMD hardware.
Ordinary consumers don't care about closed libraries any more than they cared about compilers. They care about seeing games run well on the hardware they purchase. And the problem I have with GameWorks, in a nutshell, is that it gives NV control over AMD (and Intel) GPU performance in specific areas. If NV's closed-source libraries are used in all cases, then the temptation to sabotage the competition's performance is huge.
Even if AMD can fight back by creating its own library program, it's still exacerbates a walled-garden approach that doesn't ultimately benefit the end *user.*
That's the point of the article.
The overtessellation and generally slow performance are just an example of how easy it is to create odd results. Even after analyzing the R9 290X's draw calls, it's not clear why the R9 290X is evenly matched against the GTX 770. It just is.
The point of the article is about more than overtessellation in one title. The point of the article is that closed libraries have the potential to create even more of a walled garden GPU effect.
Right now, a conventional "Gaming Evolved" or "TWIMTBP" title ships out optimized for one vendor but can still be optimized for the other post-launch. GameWorks changes that.
I consider this problematic because we've seen how companies can abuse this kind of power. 12 years ago, Intel began shipping versions of its compiler that refused to optimize for AMD hardware, even though AMD had paid Intel for the right to implement certain SIMD instruction sets. This fact wasn't widely known for years. Instead, people concluded that AMD's implementation of the various SIMD sets must have been sub-optimal, because it didn't benefit from using SSE or SSE2 the way Intel did. Since K8's implementation of SSE2 was only 64-bits wide, the conclusion was that AMD had fumbled the ball in that regard. In reality, Intel's compilers would refuse to create the most advantageous code paths for AMD hardware.
Ordinary consumers don't care about closed libraries any more than they cared about compilers. They care about seeing games run well on the hardware they purchase. And the problem I have with GameWorks, in a nutshell, is that it gives NV control over AMD (and Intel) GPU performance in specific areas. If NV's closed-source libraries are used in all cases, then the temptation to sabotage the competition's performance is huge.
Even if AMD can fight back by creating its own library program, it's still exacerbates a walled-garden approach that doesn't ultimately benefit the end *user.*
That's the point of the article.
The overtessellation and generally slow performance are just an example of how easy it is to create odd results. Even after analyzing the R9 290X's draw calls, it's not clear why the R9 290X is evenly matched against the GTX 770. It just is.
Have you heard about this guy on a forum with reading comprehension problems? Show me the link between gameworks and overtesselation and we can talk. The article doesn't do such thing. It shows that there is overtesselation, like there was in the past, when gameworks didn't exist. It also says that gameworks is closed source. It doesn't give any evidence that there's causality between them.
It's just speculation that bad things MAY happen.
Yay for clickbait!
The article is about 2 things, 1st is the closed nature of GW, the 2nd is batman tesselation tricks, they're not linked nor does the article claim they are
The latter is reported under a headline of Gameworks distorting tricks...
Well, the overtessellation part (among other examples) establishes a pattern of behavior, and GameWorks provides NVIDIA with the means to perpetrate similar acts.
firstminion
Newcomer
Precisely.
This article combines the tessellation and GameWorks discussion because that's how the story came together. When I began the investigation, I didn't know what I'd find. And I trust AMD's word on much of this partly because, in the course of working with the company, they had the opportunity to lie about smoking guns -- and didn't.
Instead of rushing to judgment with a batch of questionable data relying on old drivers and some of the initial comparisons between AMD and NV in games like Splinter Cell or Assassin's Creed IV, I took the time to chase down performance errata (Splinter Cell's patching process is basically made with wasps and sandpaper). A big expose on how NV had already crippled AMD's performance would have driven a lot more short-term traffic. But that's not what's happening here.
The takeaway isn't "AMD Good, Nvidia Bad." The takeaway is that giving Company A control over Company B's performance is never, ever a good bet for the end-user or any kind of fair competition.
Instead of rushing to judgment with a batch of questionable data relying on old drivers and some of the initial comparisons between AMD and NV in games like Splinter Cell or Assassin's Creed IV, I took the time to chase down performance errata (Splinter Cell's patching process is basically made with wasps and sandpaper). A big expose on how NV had already crippled AMD's performance would have driven a lot more short-term traffic. But that's not what's happening here.
The takeaway isn't "AMD Good, Nvidia Bad." The takeaway is that giving Company A control over Company B's performance is never, ever a good bet for the end-user or any kind of fair competition.
D
Deleted member 13524
Guest
The article talks about the nature of Gameworks and how it will further enable nVidia to use their dark-grey tactics that are not in the consumers' best interest.
Then it gives some examples of such tactics that have been used in the past as suggestion that there's a good change nVidia will do them again (spoilers: they most definitely will).
It's a "We're watching you, punk!" article with good reason to be.
I'm all about giving clicks to an article that warns consumers. I'm even willing to share it in social networks a lot more than your typical skimming-through-benchmarks reviews or pics of cute kitties.
In fact, I might even go to that page and purposedly click some ADs. That's how much I like consumer-oriented articles.
But this here is a forum for people to discuss stuff and the most effective way of calling out something as futile is to not participate in the discussion.
I think this is a desperate take on Mantle - which BTW could prove to be a lot more dangerous than TWIMTBP, GameWorks, great concrete slabs or invisible overtesselated water. If the other IHVs don't adopt it, that is.
Conclusion: library that happens to be closed source: bad. Full blown API that only works on one brand and that's closed source and proprietary: good?
Both Mantle and GameWorks have their place. But the double standard? Bah.
Both Mantle and GameWorks have their place. But the double standard? Bah.
Funny enough, Mantle could be seen like a medicine in the case a game use full gameworks ( there's not so much title yet, Splinter cell, last Assassin Creed and last batman ).. Developpers include Mantle API on side of Gameworks, like this all the problem describe in this article disappear.
The problem with Gameworks, is developpers dont let the choice, the game run with it and it seems impossible for developpers and AMD to know how it work and what performance they can have on AMD GPU ( in reality developpers could know it, at least for a specific part is to just code it normally in DX11 and compare both performance and check if it will not be a good idea to let it like that in the engine, at least on AMD GPU's ). I mean the numbers on Batman AC and Origin are funny enough.. AC was allready favor a lot Nvidia gpu's.. but when you compare how the same gpu's behave in origin and AC, the numbers one who use gameworks and one not.. you can have a good laugh. ( still the question, does it work incredbly good on a 770 or incredibly bad on a 290x )
The last problem, it seems, Nvidia have too much control on it and developpers not enough. ANd as AMD cant access the code of the game library for fix performance, bugs or even mutli-gpu support, they cant do it.. If something is broken on the day one, it will stay broken.
Dont forget Mantle is an API who work on the side of the standard engine used for the game.. Nvidia users are not forced to deal with it, and developpers will provide for nvidia users the non Mantle " version " of the game.
Last edited by a moderator:
Silent_Guy,
Let me try to explain why I see no double standard.
PhysX: Closed-source, proprietary to Nvidia. And I'm fine with that. Nvidia's PhysX program doesn't harm AMD in any fashion. It *does* give Nvidia a competitive advantage (at least in theory), but it doesn't prevent AMD from doing any form of optimization.
CUDA: Proprietary to Nvidia. I'm fine with that. Using CUDA doesn't hamper anything AMD does regarding OpenCL or DirectCompute.
Mantle: Proprietary to AMD (I don't really believe that NV can easily implement it, just as AMD couldn't implement PhysX or CUDA). Could turn out to be a competitive advantage for AMD. I'm fine with that.
GameWorks: AMD has been unable to optimize its own drivers for optimal GW performance. Tilts a previously established playing field (DX11) in a way that favors Nvidia. Explicitly prevents optimization by AMD in a manner different from typical driver optimizations.
When AMD cards had DX10.1 and NV was limited to DX10.0, that was fine. AMD had an additional functionality baked in that game developers could take advantage of. Taking advantage of DX10.1 in a game did not hurt DX10.0 implementations on NV cards.
Tilting the playing field within a given API or framework = bad. Developing new APIs and frameworks that give developers or consumers more options = not bad.
Let me try to explain why I see no double standard.
PhysX: Closed-source, proprietary to Nvidia. And I'm fine with that. Nvidia's PhysX program doesn't harm AMD in any fashion. It *does* give Nvidia a competitive advantage (at least in theory), but it doesn't prevent AMD from doing any form of optimization.
CUDA: Proprietary to Nvidia. I'm fine with that. Using CUDA doesn't hamper anything AMD does regarding OpenCL or DirectCompute.
Mantle: Proprietary to AMD (I don't really believe that NV can easily implement it, just as AMD couldn't implement PhysX or CUDA). Could turn out to be a competitive advantage for AMD. I'm fine with that.
GameWorks: AMD has been unable to optimize its own drivers for optimal GW performance. Tilts a previously established playing field (DX11) in a way that favors Nvidia. Explicitly prevents optimization by AMD in a manner different from typical driver optimizations.
When AMD cards had DX10.1 and NV was limited to DX10.0, that was fine. AMD had an additional functionality baked in that game developers could take advantage of. Taking advantage of DX10.1 in a game did not hurt DX10.0 implementations on NV cards.
Tilting the playing field within a given API or framework = bad. Developing new APIs and frameworks that give developers or consumers more options = not bad.
firstminion
Newcomer
DX: performance is in the hands of developers, AMD and nvidia.
Mantle: performance is in the hands of developers, AMD and possibly nvidia.
Gameworks: performance is in the hands of nvidia.
Reviewers don't compare Mantle performance numbers with DX because "Mantle is proprietary". Reviewers compare only DX numbers.
Gameworks substitutes DX implementations in place. Performance on both AMD and nvidia HW are in the hands of nvidia.
Gameworks poisons the tree.
AMD does the same thing with TressFX and the forward lighting Leo tech demo that was implemented in Dirt Showdown. Both vendors attempt to advance techniques that show their hardware off in the best light. It's up to game developers to choose how much they accept from AMD or Nvidia, and it's up to reviewers to decide whether it's fair to benchmark with vendor supplied code enabled or disabled (see for example the Dirt Showdown or HAWX 2 benchmarks). Then we get to discuss everything here. The more interesting things to talk about, the better IMO.
The "open" thing is a bit of a red herring, since as I've said before, vendors optimize closed code all the time via shader replacement, and both vendors work with developers to optimize game performance. I don't see how GameWorks changes much.
Finally, it's completely bizarre to me to see people saying they'd like GameWorks better if it were implemented in CUDA. I bet if Mantle wasn't looming, the people here wouldn't be half as friendly to proprietary APIs.
firstminion
Newcomer
You brought yourself the difference between gameworks and your other examples.
You can disable special effects. It may be impossible to do the same on gameworks.
That would be surprising indeed. You can turn off PhysX, Mantle, crazy tessellation in HAWX 2, CUDA effects in Just Cause 2, TressFX in Tomb Raider... I expect GameWorks will be no different.
Except that TressFX is open source, so developers who use it can see into the code, and optimize for NVIDIA if needed. I don't know about the global illumination method used in Dirt Showdown, I mean I don't know if it's open source, but I'd be surprised if Codemasters didn't have access to the code.
GameWorks's closed nature is different. In fact, if the goal really were just to improve games, why would it need to be closed? Being open-source would make it easier to reach that goal, as everyone could embrace it without having to worry about performance on Radeons, inability to fix bugs or simply tweak things.
firstminion
Newcomer
GameWorks seems to be bigger than FLEX, GI Works and Flame Works, it depends on what it targets. We don't know how big GameWorks is, that's why I believe it's better to be cautious. In comparison we know Mantle way more than Gameworks.
Its not a question of "Opensource " in the sense litteral, but in the sense of open.. if not even the developpers know what optimisation are done with nvidia hardware with it and AMD cant even optimise their driver for work on AA, SSAO, light/ shadow computing and who know what. So developpers cant optimize the game for AMD gpu's, and AMD cant even try to developp drivers for work with it. ( by working it is not even a question of "performance" but fix bugs and other things )...
Last edited by a moderator:
Similar threads
