PCPERDX12 GPU and CPU Performance Tested: Ashes of the Singularity Benchmark

[Keysplayr]

The best part is people who want to dismiss the DX12 performance, for the reasons you've sited, but want to drone on about the lack of DX11 optimizations from AMD as somehow indicative of AMD's overall DX11 performance. Then it being Alpha, a year away, not a real game, and whatever other excuses they use for nVidia's performance are not applicable for the DX11 pathway and drivers.

Droning on about the lack of DX11 optimizations from AMD as somehow indicative of AMD's overall DX11 performance.

Ummmmm, 3D, isn't that pretty much exactly what AMD proponents are trying to do here? Droning on about the good DX12 optimizations for Ashes and somehow saying it is going to be indicitive of AMD's overall DX12 performance? Why are you not upset with them. Come on man, rail into them for making those sweeping judgement calls based on this one DX12 bench. It's on you now.

 

[ocre]

Your graph is pretty useless. You could for the matter just have posted the 3Dmark API test.

So lets return to reality.

Good catch!!!!

I hadn't even noticed.

For anyone else who hasn't, take a good look at the graphs that were being posted. Misleading

 

[Piroko]

You can only log into one client at a given time. My point was there are currently more people playing FFXIV than Just Cause 2.

Did they change that? Multi-client was trivial a year ago. Also, yes, more people play MMOS than your average AAA title. But last time I checked more people played Clicker Heroes than Just Cause 2. Just goes to show where average gamers really pile up vs. where tech media covers...

You should dip into the Tera subforums. It isn't so peachy for AMD there either. Worse if you got an AMD CPU too. Woof.

Played it just fine with exactly that combination for about half a year. What I noticed most in their forums is many people who throw 180W graphics cards in old Dell econoboxes. That is bound to have issues. And AMD cards are the preferred choice for those super-budget-upgraders, due to, well, budget.

Odd, I never said or implied that. Care to show me where I mentioned that? Or are you inferring that from me stating AMD is ignoring DX11 performance in this benchmark?

See next paragraph:

Do you not see the huge performance drop in DX11 for this AMD sponsored game? They just showed us they can rekt it with the Battlefront numbers, in DX11. I can just assume AMD isn't going to touch DX11 until the end, which if that's their game plan, work it up!

Either you confused me or you just implied again that DX11 performance can only be getting worse for everything AMD because reasons. I get that you see issues in DX11 that should be resolved and I agree, but I don't see how the current situation gives a trend for the future. If anything there is this long term trend that their drivers got continuous improvements with no card-killing hiccups since 2009.

If DX12 is anything like Mantle, and it's going to require further tweaks to specific titles either via the devs or the GPU makers, there is a potential that yes future hardware might not be as optimized due these games no longer being current. But my crystal ball broke, can't say anything conclusively.

That would assume major changes in the hardware, would it not? And those can happen for both vendors. And it would assume that brute forcing DX11 is not an option - but exactly that happened with DX9 games, throw your 1,200 unified shader cores at an engine that was designed around cards with 24 unified shaders and just don't care about optimizations. But that's not specific to any vendor.

DX12 is a new animal. It's closer to the metal than PC gaming has ever been before. If the only other example is Mantle of what is probable, just look at how AMD basically ditched GCN1.2 hardware in Mantle optimizations, I mean:

http://www.anandtech.com/show/9390/the-amd-radeon-r9-fury-x-review/12

With no need to tweak Mantle for their updated GCN family of cards, that advantage to the consumer is basically gone. Where Mantle gave an edge to certain configurations that edge erased if you used a card not supported by it's older GCN build.

If the future for the API is more dev driven, we might see optimizations go the way side once game devs move on to new projects and the GPU vendors move on to new uarchs. Though of course, no crystal ball, can only go by what is the only example - and it was AMD not bothering to update Mantle for their new cards, so hello DX11 again!

Mantle gave an edge to GCN1.0 and 1.1, GCN 1.2 didn't see that. That was the situation at day one, it hasn't changed ever since. Everyone who bought a card back then knew what he got and the buyers of GCN1.2 perhaps speculated for more. That's not great, but how is this connected to DX11 and DX12?

As for you implying forced obsolescence, that fear has hit both camps in the recent past (see alledged Kepler performance degradation). Let's just see how that plays out and do some internet shitstorming whenever one of them actually tries it.

 

[railven]

Did they change that? Multi-client was trivial a year ago. Also, yes, more people play MMOS than your average AAA title. But last time I checked more people played Clicker Heroes than Just Cause 2. Just goes to show where average gamers really pile up vs. where tech media covers...

I'm confused what you mean about that. Ever since back in FFXI days you could only log into one client at a given time. If you were playing on Playstation2 you'd get booted if you logged into PC. Same with FFXIV, otherwise people would have PS3/PS4/PC clients running all in tandem. And it was all tied to one account. Ie 2.5 million subs == 2.5 unique users, but that can still translate to >2.5 million registered clients. I have the PS3/PS4/PC clients but I mostly play on PC and love the Vita feature to farm via the PS4 link.

I also feel you still miss my point. If Click Heroes was having random crashing issues on AMD hardware and had a large active user base, you don't think they'd be upset? The poster I responded to used issues with Just Cause 2 as a means to deflect the current issues FFXIV has on AMD hardware - which it for newer AMD hardware (GCN 1.2 based) it randomly crashes in DX11 mode. This doesn't affect older GCN 1.0 and 1.1 cards.

Played it just fine with exactly that combination for about half a year. What I noticed most in their forums is many people who throw 180W graphics cards in old Dell econoboxes. That is bound to have issues. And AMD cards are the preferred choice for those super-budget-upgraders, due to, well, budget.

What is fine for one, is not fine for another. But it isn't rare to read high end AMD GPUs+CPUs experiencing issues. Mostly has to do with the game's horrendous CPU bottleneck (a constant in MMOs which is a constant flaw to AMD setups. Look into Wildstar, pattern repeats. Advised to stay away from AMD CPU/GPU combinations, and the devs are blamed of being incompetent, though NV+Intel systems don't suffer as greatly. Go figure.)

See next paragraph:
Either you confused me or you just implied again that DX11 performance can only be getting worse for everything AMD because reasons. I get that you see issues in DX11 that should be resolved and I agree, but I don't see how the current situation gives a trend for the future. If anything there is this long term trend that their drivers got continuous improvements with no card-killing hiccups since 2009.

Look at the game in this example. Look at how bad the DX11 performance is in AMD hardware. I never said this was a trend, but an outlier. I even used Battlefront DX11 performance as a comparative of AMD doing DX11 right. What irks me is this game is AMD sponsored. I also said, this might not be their focus and before launch it could be rectified. As it is now, it is alarming to see AMD flounder so bad at DX11.

It could be intentional, ie "look at how much DX12 improves our performance" which can they be used as a metric for other current DX11 titles. IE "if AOTS gained almost 100% performance going from DX11 to DX12, imagine what games like Battlefront will gain when the DX12/Vulkan patch hits." And that mentality has already been seen on these forums (and the casual den I frequent - NeoGaf).

That would assume major changes in the hardware, would it not? And those can happen for both vendors. And it would assume that brute forcing DX11 is not an option - but exactly that happened with DX9 games, throw your 1,200 unified shader cores at an engine that was designed around cards with 24 unified shaders and just don't care about optimizations. But that's not specific to any vendor.

But this isn't like DX9 to DX11 where the overall API was the same. With DX12 and closer to the metal, the code has to be specifically optimized for the uarch, it seems. Anandtech threw out a few Mantle tests from their Fury X review because it performed worse than DX11. AMD said it was because they didn't optimize their Mantle for their newest uarch tweaks.

Mantle gave an edge to GCN1.0 and 1.1, GCN 1.2 didn't see that. That was the situation at day one, it hasn't changed ever since. Everyone who bought a card back then knew what he got and the buyers of GCN1.2 perhaps speculated for more. That's not great, but how is this connected to DX11 and DX12?

I'm confused on the bold. I'm not talking about DX11 and DX12 as if they are one, I'm treating them independently. IE, if DX11 is ignored (as it currently is being done in AOTS) that is a huge kick in the balls for anyone for whatever reason doesn't upgrade to Windows 10 and uses DX12.

As for you implying forced obsolescence, that fear has hit both camps in the recent past (see alledged Kepler performance degradation). Let's just see how that plays out and do some internet shitstorming whenever one of them actually tries it.

Kepler didn't lose performance. It actually gained performance. You should revisit that before you continue this argument. GCN 1.2 is over a year old, and there is still was no update for it's Mantle code. And AMD said they will no longer update it.

 

[IllogicalGlory]

For anyone else who hasn't, take a good look at the graphs that were being posted. Misleading

There was nothing misleading. The graph was clearly labeled as a draw call test, and I clearly explained why I used them (http://forums.anandtech.com/showpost.php?p=37776067&postcount=163): to prove that AMD's lack of performance in DX11 is not due to a lack of optimization, but due to a lack of peak draw calls.

I also wrote this:

I'll concede that the benchmarks don't seem to be as poor a case as that test, but it highlights exactly what's going on here. AMD's DX11 support isn't sufficient in this game, and sometimes NVIDIA's isn't either.

I'll concede that NVIDIA is playable and DX11 is fine in the actual benchmark and probably will be in gameplay. No problem there. I'll return to my original point about AMD. Without multithreaded draw calls, DX11 is a no go for them. Apart from adding that feature, there's nothing they can do for this game. Will that be a problem? Probably, but all I wanted to prove was that optimizing for DX11 is a pointless move for AMD. I think I've done that.

Come on man, really? You mean you don't see a DX11 970 score beating a DX11 980Ti score? And you don't think anything is wrong there?
Come on now.

There's no problem with that. It's margin of error. .5 FPS? What that demonstrates is that NVIDIA's performance in a draw call limited situation peaks at around 35 FPS. I can't say for sure whether draw call limitations will be a problem for them in the game, but I do concede that they're basically fine in the benchmark.

 

[Piroko]

I'm confused what you mean about that. Ever since back in FFXI days you could only log into one client at a given time. If you were playing on Playstation2 you'd get booted if you logged into PC. Same with FFXIV, otherwise people would have PS3/PS4/PC clients running all in tandem. And it was all tied to one account. Ie 2.5 million subs == 2.5 unique users, but that can still translate to >2.5 million registered clients. I have the PS3/PS4/PC clients but I mostly play on PC and love the Vita feature to farm via the PS4 link.

Oh that's the issue of understanding here. Yes, you can only log into one account from one system at a time. Switch PCs/consoles and you get booted on the other client. But you can log into several different accounts simultaneously on the same PC, let the characters sit next to each other and "chat with yourself". You'd be surprised at how many people pay for multiple subscriptions out of one reason or another in p2p MMOs. That's why numbers like 2.5 million subscriptions and 300k simultaneous logins should be taken with a truckload of salt.

I also feel you still miss my point. If Click Heroes was having random crashing issues on AMD hardware and had a large active user base, you don't think they'd be upset? The poster I responded to used issues with Just Cause 2 as a means to deflect the current issues FFXIV has on AMD hardware - which it for newer AMD hardware (GCN 1.2 based) it randomly crashes in DX11 mode. This doesn't affect older GCN 1.0 and 1.1 cards.

I agree on your point, it needs fixing. My point was more on the line that it's not perceived as a high prio issue by the press because it's not an AAA title problem. Which is stupid. Stuff like this needs press coverage to get onto priority lists.

What is fine for one, is not fine for another. But it isn't rare to read high end AMD GPUs+CPUs experiencing issues. Mostly has to do with the game's horrendous CPU bottleneck (a constant in MMOs which is a constant flaw to AMD setups. Look into Wildstar, pattern repeats. Advised to stay away from AMD CPU/GPU combinations, and the devs are blamed of being incompetent, though NV+Intel systems don't suffer as greatly. Go figure.)

Well, that was known ever since Bulldozer has been released though, so those that complain now really need to take a step back and analyze their buying pattern in my opinion. It's not something that got any worse though and with more people switching to Intel systems with higher single threaded performance is one of those issues that fixes itself, ironically speaking. It's also one of those issues that - no matter if AMD were to fix it or sit it out - would probably not increase mindshare amongst those gamers. The MMO crowd is special at times.

Look at the game in this example. Look at how bad the DX11 performance is in AMD hardware. I never said this was a trend, but an outlier. I even used Battlefront DX11 performance as a comparative of AMD doing DX11 right. What irks me is this game is AMD sponsored. I also said, this might not be their focus and before launch it could be rectified. As it is now, it is alarming to see AMD flounder so bad at DX11.

It could be intentional, ie "look at how much DX12 improves our performance" which can they be used as a metric for other current DX11 titles. IE "if AOTS gained almost 100% performance going from DX11 to DX12, imagine what games like Battlefront will gain when the DX12/Vulkan patch hits." And that mentality has already been seen on these forums (and the casual den I frequent - NeoGaf).

Well, agreed. We don't know the actual reason though and if it changes until game release, so I don't like to play the blame game yet.

But this isn't like DX9 to DX11 where the overall API was the same. With DX12 and closer to the metal, the code has to be specifically optimized for the uarch, it seems. Anandtech threw out a few Mantle tests from their Fury X review because it performed worse than DX11. AMD said it was because they didn't optimize their Mantle for their newest uarch tweaks.

So far they are optimizing for DX12 just fine though and aren't deoptimizing DX11 in the process. Nvidia is going the extra step of optimizing both for now and that should rightfully be called out in their favor.

I'm confused on the bold. I'm not talking about DX11 and DX12 as if they are one, I'm treating them independently. IE, if DX11 is ignored (as it currently is being done in AOTS) that is a huge kick in the balls for anyone for whatever reason doesn't upgrade to Windows 10 and uses DX12.

Windows 7 mainstream support has ended 7 months ago, whatever reason is one that is bound to have compromises regardless. I'm not going to say "Go 10 or go home", but I fear that's the path gamers will have to walk.

Kepler didn't lose performance. It actually gained performance. You should revisit that before you continue this argument. GCN 1.2 is over a year old, and there is still was no update for it's Mantle code. And AMD said they will no longer update it.

Alleged may have been the wrong choice of word, I'm completely aware of that.

 

[railven]

Oh that's the issue of understanding here. Yes, you can only log into one account from one system at a time. Switch PCs/consoles and you get booted on the other client. But you can log into several different accounts simultaneously on the same PC, let the characters sit next to each other and "chat with yourself". You'd be surprised at how many people pay for multiple subscriptions out of one reason or another in p2p MMOs. That's why numbers like 2.5 million subscriptions and 300k simultaneous logins should be taken with a truckload of salt.

I see where you are going with this, but I have to ask why? Why are trying to discredit the probable amount of users to this game. As if it were to make it sting less that those people (multi-boxing if you prefer) can't run the game in DX11 mode out of fear of random crashes.

I agree on your point, it needs fixing. My point was more on the line that it's not perceived as a high prio issue by the press because it's not an AAA title problem. Which is stupid. Stuff like this needs press coverage to get onto priority lists.

It is a triple AAA title, in the MMO genre. I guess WoW isn't a AAA title since it isn't covered by the major tech sites? Or any other non-benchmark laden game?

Well, that was known ever since Bulldozer has been released though, so those that complain now really need to take a step back and analyze their buying pattern in my opinion. It's not something that got any worse though and with more people switching to Intel systems with higher single threaded performance is one of those issues that fixes itself, ironically speaking. It's also one of those issues that - no matter if AMD were to fix it or sit it out - would probably not increase mindshare amongst those gamers. The MMO crowd is special at times.

Yerp, AMD's reputation is definitely in the gutter in the MMO circles I frequent.

Well, agreed. We don't know the actual reason though and if it changes until game release, so I don't like to play the blame game yet.

That's refreshing. You should look at the Ark thread, game is still in Alpha but pitchforks were raised.

So far they are optimizing for DX12 just fine though and aren't deoptimizing DX11 in the process. Nvidia is going the extra step of optimizing both for now and that should rightfully be called out in their favor.

What do you base that on? Have you not been following the recent trends? AMD is now faster than Nvidia sans GTX 980 Ti in DX11. I mean, yet again look at Battlefront. R9 390X laying the smack down on GTX 980. Yet, in AMD's sponsored game, GTX 970 is punching Fury X in the face in DX11.

http://www.computerbase.de/2015-10/star-wars-battlefront-erste-benchmarks-der-open-beta/

Windows 7 mainstream support has ended 7 months ago, whatever reason is one that is bound to have compromises regardless. I'm not going to say "Go 10 or go home", but I fear that's the path gamers will have to walk.

Again, what does this have to do with anything? You mentioned XP/Vista before, now Windows 7? I'll just assume you hated Windows 8/8.1 that much haha.

 

[IllogicalGlory]

What do you base that on? Have you not been following the recent trends? AMD is now faster than Nvidia sans GTX 980 Ti in DX11. I mean, yet again look at Battlefront. R9 390X laying the smack down on GTX 980. Yet, in AMD's sponsored game, GTX 970 is punching Fury X in the face in DX11.

It's not a lack of miscellaneous optimizations, it's that their single threaded draw calls can't handle the amount that that game puts out. The only solution to this is achieving draw call parity with NVIDIA, which they obviously aren't capable of doing. They are behind in Battlefront as well, at least on i3s. Battlefront with an AMD card and i3 is a disaster.

Again, what does this have to do with anything? You mentioned XP/Vista before, now Windows 7? I'll just assume you hated Windows 8/8.1 that much haha.

Are you confusing him with me?

People who want to stick with 7/8 will eventually be forced to change OSs. You don't see anyone gaming on Windows Vista these days; even if they don't like MS's practices with 10, if even half of what we've heard about DX12 comes true, but they'll be in a very difficult position when they're forced to shell out $100 to a company they're protesting against. That or they'll have to play new games on the PS4.

 

[bystander36]

There was nothing misleading. The graph was clearly labeled as a draw call test, and I clearly explained why I used them (http://forums.anandtech.com/showpost.php?p=37776067&postcount=163): to prove that AMD's lack of performance in DX11 is not due to a lack of optimization, but due to a lack of peak draw calls.

You did fail to make it known. At least I didn't see it, and most people here didn't either.

The other point is that it doesn't proove that draw calls are the reason for the problem. Just that there is a draw call limit that is lower, but it is rare that games push up against the draw call limit, except in short spurts. It's also quite possible AMD could spend time on their drivers to improve that issue on the game.

I find it hard to believe that only games that have Mantle and alpha/beta games with DX12 have draw call limits that prevent them from competing with Nvidia in DX11. Pretty much all other DX11 games show comparable results.

 

[IllogicalGlory]

You did fail to make it known. At least I didn't see it, and most people here didn't either.

The other point is that it doesn't proove that draw calls are the reason for the problem. Just that there is a draw call limit that is lower, but it is rare that games push up against the draw call limit, except in short spurts. It's also quite possible AMD could spend time on their drivers to improve that issue on the game.

I find it hard to believe that only games that have Mantle and alpha/beta games with DX12 have draw call limits that prevent them from competing with Nvidia in DX11. Pretty much all other DX11 games show comparable results.

The graph is clearly labeled. Other language or not, what it's testing is made clear. Draw calls.

Take a look at them again. If the difference between AMD and NVIDIA were simply DX11 optimizations, how do you explain the results?

Normal draw calls

970 is 3% faster than 390.

High draw calls:

970 is 67% faster than the 390.

If the difference is regular DX11 optimizations, why are these results so different?

I am not saying that all DX12/Mantle games have draw call limitations, but it's clear looking at this particular game that it's using DX12 chiefly for the advantages in draw calls. It's supposed to deliver massive battles with all kinds of stuff on the screen. Have you seen the benchmark run?

 

[bystander36]

The graph is clearly labeled. Other language or not, what it's testing is made clear. Draw calls.

Take a look at them again. If the difference between AMD and NVIDIA were simply DX11 optimizations, how do you explain the results?

Normal draw calls

970 is 3% faster than 390.



High draw calls:

970 is 67% faster than the 390.



If the difference is regular DX11 optimizations, why are these results so different?

I am not saying that all DX12/Mantle games have draw call limitations, but it's clear looking at this particular game that it's using DX12 chiefly for the advantages in draw calls. It's supposed to deliver massive battles with all kinds of stuff on the screen. Have you seen the benchmark run?

As far as the graph being labled goes, it is, but most people see a graph assume it's a FPS graph and don't look for the that information.

And I did already give a reasonable explination. Driver optimizations, and possibly even AMD helping with game optimizations.

Don't you find it odd that ONLY Mantle/DX 12 games have AMD performance so far behind Nvidia's in DX11 (ignoring Gameworks too, but that's another issue)? If this was a wide spread issue, you'd see problems in DX11 games that don't offer Mantle/DX12.

You've already seen what optimizations can do between now and the previous version of the game and drivers. It may require effort, but it can be done.

 

[IllogicalGlory]

Don't you find it odd that ONLY Mantle/DX 12 games have AMD performance so far behind Nvidia's in DX11 (ignoring Gameworks too, but that's another issue)? If this was a wide spread issue, you'd see problems in DX11 games that don't offer Mantle/DX12.

Which games? AMD performs fine in DX11 with Thief, Battlefield 4 and DA:I. Fable Legends doesn't seem to have a DX11 path, at least for the benchmark.

If it's Star Swarm or something like that, the draw calls come into play with this game because it's a design choice. You don't see this problem in other DX11 titles because they know that draw calls are limited and aim not to exceed that. This DX12 game doesn't have to worry about it, so you're seeing AMD's lack of draw call capacity coming into play.

You've already seen what optimizations can do between now and the previous version of the game and drivers. It may require effort, but it can be done.

All I've seen is DX12 performance improving. DX11 performance hasn't changed that much, even for NVIDIA. As long as the draw call limit isn't being hit, normal game-specific optimizations will work, but AMD has this really bizarre discrepancy that only exists in this game. There's only one thing about this game that's special as far as I can tell.

 

[bystander36]

Which games? AMD performs fine in DX11 with Thief, Battlefield 4 and DA:I. Fable Legends doesn't seem to have a DX11 path, at least for the benchmark, which games are you talking about?

If it's Star Swarm or something like that, those are designed to be draw call limited as well. The draw calls come into play with this game because it's a design choice. You don't see this problem in other DX11 titles because they know that draw calls are limited and aim not to exceed that. This DX12 game doesn't have to worry on that front.All I've seen is DX12 performance improving. DX11 performance hasn't changed that much, even for NVIDIA.

Those games at released struggled in DX11 mode. They've worked it out now, but at release, you pretty much had to go Mantle for good performance.

There is a good chance they'll eventually get the DX11 path in this game working too. It just isn't there now.

 

[IllogicalGlory]

I don't see that.

All of these are from the first week of release:

http://www.techspot.com/review/734-battlefield-4-benchmarks/
http://www.techspot.com/review/787-thief-benchmarks/page2.html
http://www.techspot.com/review/921-dragon-age-inquisition-benchmarks/page4.html

If anything, AMD seems to be doing too well.

 

[bystander36]

I don't see that.

All of these are from the first week of release:

http://www.techspot.com/review/734-battlefield-4-benchmarks/
http://www.techspot.com/review/787-thief-benchmarks/page2.html
http://www.techspot.com/review/921-dragon-age-inquisition-benchmarks/page4.html







If anything, AMD seems to be doing too well.

Then that should tell you that this game should work well in DX11, as I know DA:I for example, is very draw call limited.

I must be recalling Beta benchmarks on those games.

There is no reason for AMD to be this far behind on DX11, other than a lack of attention.

 

[3DVagabond]

It's not a lack of miscellaneous optimizations, it's that their single threaded draw calls can't handle the amount that that game puts out. The only solution to this is achieving draw call parity with NVIDIA, which they obviously aren't capable of doing. They are behind in Battlefront as well, at least on i3s. Battlefront with an AMD card and i3 is a disaster.Are you confusing him with me?

People who want to stick with 7/8 will eventually be forced to change OSs. You don't see anyone gaming on Windows Vista these days; even if they don't like MS's practices with 10, if even half of what we've heard about DX12 comes true, but they'll be in a very difficult position when they're forced to shell out $100 to a company they're protesting against. That or they'll have to play new games on the PS4.

I don't think it's an issue of not being capable. They just feel that the gains to be had are too minimal and instead DX11 needs to be replaced. That's why they came out with Mantle. I'm sure if they could manage Mantle they'd have little issues with drivers.

 

[ThatBuzzkiller]

People here forget that that Nvidia's front end is stronger than AMD's when it comes to pushing realistic geometry throughput ...

Nvidia can afford to do less batching therefore they can submit less triangles per draw to get maximum performance whereas AMD needs to batch more triangles per draw ...

A part of AMD's weak showing in some games with DX11 mode is attributed to some hardware limitations, not a neglect of driver optimizations ...

 

[Piroko]

I see where you are going with this, but I have to ask why? Why are trying to discredit the probable amount of users to this game. As if it were to make it sting less that those people (multi-boxing if you prefer) can't run the game in DX11 mode out of fear of random crashes.

It is a triple AAA title, in the MMO genre. I guess WoW isn't a AAA title since it isn't covered by the major tech sites? Or any other non-benchmark laden game?

No discrediting anything, though you started the numbers game and I gave some critique that numbers aren't neccessarily comparable between genres :awe:

And it's definitely not covered as a AAA title in western press, the amount of random coverage even the much older WOW gets is ridiculous in comparison, so there's that. You naming those issues was the first time I ever heard about them and the first time that I read anyting about FF14 in 6 months...

That's refreshing. You should look at the Ark thread, game is still in Alpha but pitchforks were raised.

At least that's a game one can play, uninteresting as it is. I won't complain over raised pitchforks of Ark issues, though I might question the target in that specific case.

What do you base that on? Have you not been following the recent trends? AMD is now faster than Nvidia sans GTX 980 Ti in DX11. I mean, yet again look at Battlefront. R9 390X laying the smack down on GTX 980. Yet, in AMD's sponsored game, GTX 970 is punching Fury X in the face in DX11.

http://www.computerbase.de/2015-10/star-wars-battlefront-erste-benchmarks-der-open-beta/

Personally, I'd say Battlefront is as much of an outlier as Ashes is, it's one of those few games that run great even if you start it the first time and sail away with default settings. Really liked that part of it.

Again, what does this have to do with anything? You mentioned XP/Vista before, now Windows 7? I'll just assume you hated Windows 8/8.1 that much haha.

I had Win2k for the longest time (never change a running system, especially when the windows install makes it through two complete hardware changes without hiccups), switched to Vista when it came out (really liked it back then actually, had very few issues with it except with some annoyances with both Creative and Nvidia drivers) and then Win 8 when that came out. Also really liked that, especially with a touchscreen device

I named Win7 due to install base, its age and the amount of - imho, unjustified - fanboyism that surrounds 7.

 

[ShintaiDK]

Guess someone lacks money. AOTS is now early access on steam:
http://store.steampowered.com/app/228880/

20% discount for the time being. Game is a year away or more.

 

[tential]

Guess someone lacks money. AOTS is now early access on steam:
http://store.steampowered.com/app/228880/

20% discount for the time being. Game is a year away or more.

I feel like early access is just another way to pre-order now. So I guess we should know see a lot more benches?

 

[guskline]

Guess someone lacks money. AOTS is now early access on steam:
http://store.steampowered.com/app/228880/

20% discount for the time being. Game is a year away or more.

I bought it to learn AND to run the benchmarks on my rigs below.:biggrin:

 

[Jaydip]

Gamegpu Test

http://gamegpu.ru/rts-/-strategii/ashes-of-the-singularity-test-gpu.html

It seems 980Ti is leading Fury X in every situation now and sometimes very convincingly, I am glad that the 980Ti is bad at Dx12 myth is busted for good.

 

[desprado]

Gamegpu Test

http://gamegpu.ru/rts-/-strategii/ashes-of-the-singularity-test-gpu.html

It seems 980Ti is leading Fury X in every situation now and sometimes very convincingly, I am glad that the 980Ti is bad at Dx12 myth is busted for good.

There is no excuse now to get Fury X over GTX 980 Ti.

 

[Goatsecks]

Gamegpu Test

http://gamegpu.ru/rts-/-strategii/ashes-of-the-singularity-test-gpu.html

It seems 980Ti is leading Fury X in every situation now and sometimes very convincingly, I am glad that the 980Ti is bad at Dx12 myth is busted for good.

My prediction: the data will be ignored or rubbished.

 

[Techhog]

Gamegpu Test

http://gamegpu.ru/rts-/-strategii/ashes-of-the-singularity-test-gpu.html

It seems 980Ti is leading Fury X in every situation now and sometimes very convincingly, I am glad that the 980Ti is bad at Dx12 myth is busted for good.

Godspeed, AMD. Hopefully you find a way to go out with dignity.