CIV: BE benchmarked

Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

KaRLiToS

Golden Member
Jul 30, 2010
1,918
11
81
So apparently to bench the game you need to add this at the end of the target location: "-benchmark lategameview"

"E: \SteamLibrary\SteamApps\Common\Sid Meiers Civilization Beyond Earth\CivilizationBe_DX11.exe" -benchmark lategameview

(or add the -benchmark lategameview in the Launch options in Steam)

14bp64i.jpg


It will give you a .cvs file name lategameviewDX.csv or lategameviewMANTLE.csv

Looking like this, (please don't use that file as an example, it was run with 14.9.1 with Ultra settings at 8040x1440 with only single card since we don't have 14.9.2)

Code:
0.000000,169.995392,170.373077,321.533020,152.455322,140.268723,155.000900,182.420837,172.113281,171.472687,158.039490,167.830231,170.461334,183.934860,178.956375,178.261246,162.546341,149.139053,147.275650,152.120117,155.135818,305.416046,138.471283,152.128540,156.317947,160.394409,154.298843,162.771011,155.123779,128.342621,155.028000,158.350006,134.396332,153.553421,146.188705,156.670334,140.418701,158.235565,127.850502,142.220978,157.518143,310.225586,127.512589,160.128769,157.110657,122.271164,142.454681,159.479126,159.864639,145.224319,157.396484,154.213623,159.348114,174.286896,143.395264,146.537766,156.122498,156.900726,158.728287,154.618698,128.886261,145.025238,137.270798,161.597626,162.559891,156.533600,311.129120,163.453491,149.792313,174.484772,152.999863,146.049561,180.463776,174.346222,153.873581,150.552185,170.773041,164.219696,178.518158,174.095047,153.288391,150.413635,168.516907,157.185638,315.406158,163.249893,151.849960,154.105194,144.322586,176.493942,166.218307,167.322739,150.088669,171.753677,169.153900,170.156830,171.321793,166.233673,146.109787,177.537537,170.186661,142.070984,295.556641,174.152267,161.509384,166.471298,184.930557,180.674301,181.226974,168.616592,175.784363,183.543045,188.457672,162.010834,166.820358,182.519928,183.651459,195.027588,335.339996,204.499374,176.324371,202.938660,202.825119,198.183960,200.917755,182.321762,209.407394,208.818588,190.159042,214.570511,190.268661,219.854675,235.573212,383.757843,189.086227,192.494980,203.406693,232.233444,216.746216,199.799774,224.182632,223.543518,227.084488,230.802551,214.305161,215.863464,213.471207,199.999771,217.502472,234.520004,228.062119,352.706268,198.422485,224.363342,226.106567,187.660156,227.107986,207.870773,234.512772,234.222122,226.795059,225.777069,204.068390,232.628281,359.369537,208.776718,210.419952,226.257751,230.457397,232.016907,218.984283,231.850647,231.610306,231.801254,200.296112,194.930313,218.237061,372.248596,230.022186,197.357819,228.648514,228.124451,233.084274,231.875351,185.459442,219.467377,213.700989,217.628677,230.002914,225.976151,226.209869,378.820587,220.437469,215.268326,207.597000,227.446518,233.720673,178.549484,223.280914,208.265015,206.181473,206.879593,164.502792,210.640411,191.849243,184.292969,169.871323,188.448044,187.056885,190.333405,341.773468,177.056244,173.881210,185.931976,182.011841,182.846100,169.290939,144.968628,144.658401,170.015869,170.560410,168.702133,168.035324,169.269867,150.481705,138.381226,152.445999,172.616257,324.532196,147.743088,149.649551,173.934525,170.681183,169.836075,169.785172,169.115967,156.476379,167.257385,161.622620,179.005768,169.142151,168.546127,129.979233,149.375473,175.934052,166.111984,173.327637,181.666382,164.391357,159.422821,174.623627,323.937073,176.106934,167.632950,153.299240,147.210297,159.441177,159.882706,159.896255,164.185364,151.357834,151.315369,164.468460,157.963593,159.692657,161.899994,154.466309,148.830338,


__________________________

I also tried to use Optimize 1x1 as CrossFire Mode but the benchmark was skipping and stuttering (Giga Stuttering) and still only one card was utilized.
 
Last edited:

el etro

Golden Member
Jul 21, 2013
1,581
14
81
If this turns into an AMD vs. NVidia war with price bashing over a single game benchmark, I will be closing this thread and infracting/warning everyone above that took the bait.

From my post forward, get this discussion back on track, now.

-Moderator Rvenger

One guy dropped the bomb on this thread. You know who is he. Punish him first.


Warning issued for moderator callout. Next time just report the post.

-Rvenger
 
Last edited by a moderator:

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
R9 290X is now $300 and 290 drops to $250-275 regularly. Having said that 980 was a poor value even without 290 cards due to the 970. The scaling 290X produces over my 7970ghz is impressive in Ryse: Son of Rome and this game, which goes to show most games don't optimize well enough for GCN. Now that AMD has already doubled the ROPs in Hawaii, I am hoping 390X they focus on adding more stream processors instead, say 60 CUs or 3840 SPs. That card would nearly double mine in performance and should be competitive with GM200. For that reason if AMD is waiting for 20nm, I think that's a smart decision. Hopefully more games take advantage of Mantle like CIV: BE.

The 970 or 290 weren't on that chart. It was the 980 and the 290X. Thus my comparison.

Besides, nVidia having some other card than the 980 being a better deal doesn't excuse the 980.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
These benchmarks do kind of make it seem like Mantle is just AMD's equivalent of DX11 multithreading. Depending on which one is more widely supported, I'd rather have that vendor's card(s).

Look at the minimums. It's more than just multithreading. It actually reduces CPU overhead.
 

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
Where's PCPer's FCAT results? I want to see the "observed FPS" not the reported that counts dropped frames.

With FCAT they've got mother of all benchmarking tools, but everyone is using it as they see fit.
There are no common standard on its use.

With Civ BE we are even seeing comeback of "minimums".
Really? Minimum fps in 2014?

With images like this dafuq would I need "minimum" fps that carries no info on minimum longevity.
be-25.png


Where are 1% and 99%. Where is frametime variation?
Instead we are getting "minimums" LMAO
 

RaulF

Senior member
Jan 18, 2008
844
1
81
So apparently to bench the game you need to add this at the end of the target location: "-benchmark lategameview"

"E: \SteamLibrary\SteamApps\Common\Sid Meiers Civilization Beyond Earth\CivilizationBe_DX11.exe" -benchmark lategameview

(or add the -benchmark lategameview in the Launch options in Steam)

14bp64i.jpg


It will give you a .cvs file name lategameviewDX.csv or lategameviewMANTLE.csv

Looking like this, (please don't use that file as an example, it was run with 14.9.1 with Ultra settings at 8040x1440 with only single card since we don't have 14.9.2)

Code:
0.000000,169.995392,170.373077,321.533020,152.455322,140.268723,155.000900,182.420837,172.113281,171.472687,158.039490,167.830231,170.461334,183.934860,178.956375,178.261246,162.546341,149.139053,147.275650,152.120117,155.135818,305.416046,138.471283,152.128540,156.317947,160.394409,154.298843,162.771011,155.123779,128.342621,155.028000,158.350006,134.396332,153.553421,146.188705,156.670334,140.418701,158.235565,127.850502,142.220978,157.518143,310.225586,127.512589,160.128769,157.110657,122.271164,142.454681,159.479126,159.864639,145.224319,157.396484,154.213623,159.348114,174.286896,143.395264,146.537766,156.122498,156.900726,158.728287,154.618698,128.886261,145.025238,137.270798,161.597626,162.559891,156.533600,311.129120,163.453491,149.792313,174.484772,152.999863,146.049561,180.463776,174.346222,153.873581,150.552185,170.773041,164.219696,178.518158,174.095047,153.288391,150.413635,168.516907,157.185638,315.406158,163.249893,151.849960,154.105194,144.322586,176.493942,166.218307,167.322739,150.088669,171.753677,169.153900,170.156830,171.321793,166.233673,146.109787,177.537537,170.186661,142.070984,295.556641,174.152267,161.509384,166.471298,184.930557,180.674301,181.226974,168.616592,175.784363,183.543045,188.457672,162.010834,166.820358,182.519928,183.651459,195.027588,335.339996,204.499374,176.324371,202.938660,202.825119,198.183960,200.917755,182.321762,209.407394,208.818588,190.159042,214.570511,190.268661,219.854675,235.573212,383.757843,189.086227,192.494980,203.406693,232.233444,216.746216,199.799774,224.182632,223.543518,227.084488,230.802551,214.305161,215.863464,213.471207,199.999771,217.502472,234.520004,228.062119,352.706268,198.422485,224.363342,226.106567,187.660156,227.107986,207.870773,234.512772,234.222122,226.795059,225.777069,204.068390,232.628281,359.369537,208.776718,210.419952,226.257751,230.457397,232.016907,218.984283,231.850647,231.610306,231.801254,200.296112,194.930313,218.237061,372.248596,230.022186,197.357819,228.648514,228.124451,233.084274,231.875351,185.459442,219.467377,213.700989,217.628677,230.002914,225.976151,226.209869,378.820587,220.437469,215.268326,207.597000,227.446518,233.720673,178.549484,223.280914,208.265015,206.181473,206.879593,164.502792,210.640411,191.849243,184.292969,169.871323,188.448044,187.056885,190.333405,341.773468,177.056244,173.881210,185.931976,182.011841,182.846100,169.290939,144.968628,144.658401,170.015869,170.560410,168.702133,168.035324,169.269867,150.481705,138.381226,152.445999,172.616257,324.532196,147.743088,149.649551,173.934525,170.681183,169.836075,169.785172,169.115967,156.476379,167.257385,161.622620,179.005768,169.142151,168.546127,129.979233,149.375473,175.934052,166.111984,173.327637,181.666382,164.391357,159.422821,174.623627,323.937073,176.106934,167.632950,153.299240,147.210297,159.441177,159.882706,159.896255,164.185364,151.357834,151.315369,164.468460,157.963593,159.692657,161.899994,154.466309,148.830338,


__________________________

I also tried to use Optimize 1x1 as CrossFire Mode but the benchmark was skipping and stuttering (Giga Stuttering) and still only one card was utilized.

Hows the bench marking coming along.

Really interested on 2 gpu croosfire.

Thanks.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
What are you talking?
This here is about CIV and not some other games. And nVidia cards have a 50% advantage with DX.

And with Mantle there is literally no improvement over a single card and with DX you are limited by the dx driver overhead.

I am talking about facts based on professional reviews. In your world NV's 980 is 50% faster than 290X in DX? :sneaky:

68436.png


68437.png


You also stated Mantle provides no advantage for single graphics card over DX completely missing the major improvement in minimum frames per second that Mantle brings.

68441.png


68442.png


Since Mantle doesn't produce any artifacts in this game, why wouldn't a gamer with a GCN GPU enable Mantle for single graphics card system? They would.

It's highly probable that AMD worked closely with the developer to help them optimize the game for Mantle instead of DX11. Making statements like NV is 50% faster in DX11 misses this point completely, not to mention the statement itself is pure fabrication of how the game performs in DX11 on AMD hardware.

Your statements are also contradictory in nature. You first stated that NV is 50% faster in DX11, yet you say Mantle brings no improvement. That would suggest that NV in DX11 would also be 50% faster than AMD with Mantle. Again, something you just made up out of thin air by ignoring 99% of reviews and cherry-picking some site that has consistently produced the most ludicrous data. :thumbsdown:
 

SimianR

Senior member
Mar 10, 2011
609
16
81
With FCAT they've got mother of all benchmarking tools, but everyone is using it as they see fit.
There are no common standard on its use.

With Civ BE we are even seeing comeback of "minimums".
Really? Minimum fps in 2014?

With images like this dafuq would I need "minimum" fps that carries no info on minimum longevity.


Where are 1% and 99%. Where is frametime variation?
Instead we are getting "minimums" LMAO

Because Mantle is a new API - and it's worth investigating all the ways that it impacts performance. Looking at average fps and throwing up your hands and saying "I'm done" is what an average reviewer would do. But I'm sorry - if NVIDIA cards are dropping to 30 fps at certain points and the AMD cards aren't dropping below 50, you don't think that's relevant? You don't think that's important to make note of?

I'm also noticing a lot of people mentioning the poor performance of AMD cards under DX - but frankly why would AMD or Firaxis waste their time getting these cards performing well under DX11 in this game in particular if no one in their right mind would run the game under that API? Not to mention that the "poor dx11 performance" of AMD cards seem to be restricted to one site. To me criticizing DX11 performance and completely ignoring the performance gains of and improvements of Mantle really is just grasping at straws and looking to make the situation look negative.
 
Last edited:

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
Because Mantle is a new API - and it's worth investigating all the ways that it impacts performance. Looking at average fps and throwing up your hands and saying "I'm done" is what an average reviewer would do. But I'm sorry - if NVIDIA cards are dropping to 30 fps at certain points and the AMD cards aren't dropping below 50, you don't think that's relevant?

Please re-read what I wrote. That's the last thing I would suggest.

And I wasn't commenting on AMD vs NV Civ:BE results. Games come and go.
I was more concerned with benchmarking methodology that this time around seems to be centered on "minimums"

But if you really want:
After browsing through several benchmarks it's evident that NV's DX works almost as good as Mantle; ofc not quite there.
It's AMD's DX that is in another low-league all together
 

SimianR

Senior member
Mar 10, 2011
609
16
81
After browsing through several benchmarks it's evident that NV's DX works almost as good as Mantle; ofc not quite there.
It's AMD's DX that is in another low-league all together
Which ones? Because the majority of reputable sites are showing the AMD DX11 benchmarks slightly behind and the Mantle benchmarks slightly ahead, with drastically different minimum fps. I'm not sure why you think a 20fps minimum fps difference wouldn't be noteworthy.

Let me ask you another question: Do you think the DX11 path in Civ: Beyond Earth is representative of AMD's DX11 performance in all games? Is it really that far fetched to believe that optimization was limited under DX11 in a title that they believed no one would bother using when a superior performance option (mantle) is available?
 
Last edited:

Tattare

Senior member
Mar 4, 2008
231
0
76
im noticing a big drop in fps running mantle but also 14.9 drivers. 40fps vs 29fps in 1920x1080, 8x, ultra
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
But if you really want:
After browsing through several benchmarks it's evident that NV's DX works almost as good as Mantle; ofc not quite there

I wouldn't say that a $550 card barely beating a $300 R9 290X or 970 losing to R9 290 is "almost as good as Mantle." Also, if you look across a spectrum of sites that tested the game, DX11 performance is still very good, not that it matters since no one with a single GPU and AMD card will use DX11 when Mantle is simply superior. AMD just didn't optimize their DX11 driver for this game since what's the point of wasting their financial resources on 2 different APIs?! It's like would you run the game in DX if OpenGL produced better performance with no adverse side-effects? No, you wouldn't. It should be a matter of time before AMD releases a driver that enables CF support under Mantle.


Warning issued for continuing a pricing debate. Did you not see my warning?? http://forums.anandtech.com/showpost.php?p=36839598&postcount=94

-Moderator Rvenger
 
Last edited by a moderator:

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
Which ones? Because the majority of reputable sites are showing the AMD DX11 benchmarks slightly behind and the Mantle benchmarks slightly ahead, with drastically different minimum fps. I'm not sure why you think a 20fps minimum fps difference wouldn't be noteworthy.

Let me ask you another question: Do you think the DX11 path in Civ: Beyond Earth is representative of AMD's DX11 performance in all games? Is it really that far fetched to believe that optimization was limited under DX11 in a title that they believed no one would bother using when a superior performance option (mantle) is available?

This image answers all your questions and then some.

http://www.pcper.com/files/imagecache/article_max_width/review/2014-10-23/be-25.png

Basically we have 290 Mantle >= 980 >> 290 DX.

Ofc someone looking at the minimum of the same run would see this as

290 Mantle >> 980 > 290 DX which is simply not true looking at the all-encompassing-truth which is frametime graph.

Minimums in 2014, after all those frame-times and variances discussions is a step backwards.
Proly inspired by AMD benchmarking guidelines.

Note that frametime graph, 99th percentile frametime or 99.9%, all these methods capture minimum fps behaviour and unlike "minimums" can't be fooled with examples like the one above.
 

DooKey

Golden Member
Nov 9, 2005
1,811
458
136
I wouldn't say that a $550 card barely beating a $300 R9 290X or 970 losing to R9 290 is "almost as good as Mantle." Also, if you look across a spectrum of sites that tested the game, DX11 performance is still very good, not that it matters since no one with a single GPU and AMD card will use DX11 when Mantle is simply superior. AMD just didn't optimize their DX11 driver for this game since what's the point of wasting their financial resources on 2 different APIs?! It's like would you run the game in DX if OpenGL produced better performance with no adverse side-effects? No, you wouldn't. It should be a matter of time before AMD releases a driver that enables CF support under Mantle.

If AMD continues to neglect DX11 as you infer then that isn't a good model for them to pursue. If all of their GPUs could use Mantle then no big deal, but unfortunately for them they can't. You are flat out wrong in your silly conjecture.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,107
1,260
126
I wouldn't say that a $550 card barely beating a $300 R9 290X or 970 losing to R9 290 is "almost as good as Mantle." Also, if you look across a spectrum of sites that tested the game, DX11 performance is still very good, not that it matters since no one with a single GPU and AMD card will use DX11 when Mantle is simply superior. AMD just didn't optimize their DX11 driver for this game since what's the point of wasting their financial resources on 2 different APIs?! It's like would you run the game in DX if OpenGL produced better performance with no adverse side-effects? No, you wouldn't. It should be a matter of time before AMD releases a driver that enables CF support under Mantle.

Yeah Mantle is really impressive in this game. There was a lot of expectation that this sort of genre would see better improvements than something like BF4 did.

Mantle delivers in this title big time. Those minimum frame improvements are exceptional. AMD should now be pushing anyone who makes RTS games to get the Mantle path added. It would totally change the performance landscape in games like Starcraft 2 that bog down no matter how strong your GPUs are because of CPU bottlenecks.
 

SimianR

Senior member
Mar 10, 2011
609
16
81
This image answers all your questions and then some.

http://www.pcper.com/files/imagecach...0-23/be-25.png

Basically we have 290 Mantle >= 980 >> 290 DX.

Ofc someone looking at the minimum of the same run would see this as

290 Mantle >> 980 > 290 DX which is simply not true looking at the all-encompassing-truth which is frametime graph.

Minimums in 2014, after all those frame-times and variances discussions is a step backwards.
Proly inspired by AMD benchmarking guidelines.

Note that frametime graph, 99th percentile frametime or 99.9%, all these methods capture minimum fps behaviour and unlike "minimums" can't be fooled with examples like the one above.
So frame time is the only important metric now? Why can't minimums be a part of the discussion as well? To me they're just as important in 2014 as they were in 2004. I'm not sure why you seem to be focused on one measurement as the sole indicator of performance overall.

If AMD continues to neglect DX11 as you infer then that isn't a good model for them to pursue. If all of their GPUs could use Mantle then no big deal, but unfortunately for them they can't. You are flat out wrong in your silly conjecture.
I think it's silly to assume that AMD would neglect DX11 as a whole, especially considering not all games support Mantle. But I think it's also silly to assume that AMD would spend optimizing for DX11 performance in a game that supports Mantle.

I think the fact is that the 980 is a faster card, from the ground up, than the 290X. It's mantle that's allowing AMD to compete and actually overtake NVIDIA's GPU's in this game. Instead, it seems people are looking at this as AMD playing catch-up with their "poor" DX11 support, but really it's impressive that AMD can be on par with and surpass NVIDIA's flagship. I guess it's all perspective.
 
Last edited:

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
So no, I am not against minimums per se, and I am not against recording minimum fps behaviour.

Bare minimums, a single number which can be meaningless(or not) is what I am against.
 

f1sherman

Platinum Member
Apr 5, 2011
2,243
1
0
So frame time is the only important metric now? Why can't minimums be a part of the discussion as well? To me they're just as important in 2014 as they were in 2004. I'm not sure why you seem to be focused on one measurement as the sole indicator of performance overall.

JHC.... because
  • bare minimum is incorporated in frame time graph
  • because minimum fps behaviour is more thoroughly described with more advanced metrics like 99th and 99.9 percentile
  • bare minimum can be deceitful as in the example above

And yes frame times is the only important metric. Allways been. It describes everything perf. related except (input) latency.
(Bare) minimum is a joke metrics
 
Last edited:

Grooveriding

Diamond Member
Dec 25, 2008
9,107
1,260
126
So frame time is the only important metric now? Why can't minimums be a part of the discussion as well? To me they're just as important in 2014 as they were in 2004. I'm not sure why you seem to be focused on one measurement as the sole indicator of performance overall.

I think it's silly to assume that AMD would neglect DX11 as a whole, especially considering not all games support Mantle. But I think it's also silly to assume that AMD would spend optimizing for DX11 performance in a game that supports Mantle.

I think the fact is that the 980 is a faster card, from the ground up, than the 290X. It's mantle that's allowing AMD to compete and actually overtake NVIDIA's GPU's in this game. Instead, it seems people are looking at this as AMD playing catch-up with their "poor" DX11 support, but really it's impressive that AMD can be on par with and surpass NVIDIA's flagship. I guess it's all perspective.


Quite true. Mantle is delivering in this title and minimums are one of the most important metrics. When you're playing a game cruising along at a good framerate and it drops suddenly to a much lower rate due to increased activity, it's one of the most annoying things if you care about the visual experience. This is why I personally go overkill on GPUs for my resolution, to avoid exactly this happening.

I think what you're seeing is just an age old AT VC&G deflection known as shifting the goalposts :D
 

Paul98

Diamond Member
Jan 31, 2010
3,732
199
106
JHC.... because
  • bare minimum is incorporated in frame time graph
  • because minimum fps behaviour is more thoroughly described with more advanced metrics like 99th and 99.9 percentile
  • bare minimum can be deceitful as in the example above

And yes frame times is the only important metric. Allways been. It describes everything perf. related except (input) latency.
(Bare) minimum is a joke metrics

It's not a joke metric as if you get a certain minimum above what you find is reasonable you will get good performance. But that isn't to say that a lower minimum can't also be very good performance wise if that minimum is only hit very rarely or when it doesn't matter.
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
I am talking about facts based on professional reviews. In your world NV's 980 is 50% faster than 290X in DX? :sneaky:

Since Mantle doesn't produce any artifacts in this game, why wouldn't a gamer with a GCN GPU enable Mantle for single graphics card system? They would.

It's highly probable that AMD worked closely with the developer to help them optimize the game for Mantle instead of DX11. Making statements like NV is 50% faster in DX11 misses this point completely, not to mention the statement itself is pure fabrication of how the game performs in DX11 on AMD hardware.

Your statements are also contradictory in nature. You first stated that NV is 50% faster in DX11, yet you say Mantle brings no improvement. That would suggest that NV in DX11 would also be 50% faster than AMD with Mantle. Again, something you just made up out of thin air by ignoring 99% of reviews and cherry-picking some site that has consistently produced the most ludicrous data. :thumbsdown:

Don't take him out of context. He was talking about drivers and you bring in 4K and 1400p benchmarks. He meant improvement over Nvidia's DX.

Which ones? Because the majority of reputable sites are showing the AMD DX11 benchmarks slightly behind and the Mantle benchmarks slightly ahead, with drastically different minimum fps. I'm not sure why you think a 20fps minimum fps difference wouldn't be noteworthy.

Let me ask you another question: Do you think the DX11 path in Civ: Beyond Earth is representative of AMD's DX11 performance in all games? Is it really that far fetched to believe that optimization was limited under DX11 in a title that they believed no one would bother using when a superior performance option (mantle) is available?

Well, all those who bought and are still on 6xxx series then got shafted. And yes, if you look at many other games (non-mantle) you generally see that nvidia performs with less overhead in DX 11.

Personally, Nvidia's DX and AMD's mantle seem to work pretty similarly in this game with AMD pulling ahead.

I wouldn't say that a $550 card barely beating a $300 R9 290X or 970 losing to R9 290 is "almost as good as Mantle." Also, if you look across a spectrum of sites that tested the game, DX11 performance is still very good, not that it matters since no one with a single GPU and AMD card will use DX11 when Mantle is simply superior. AMD just didn't optimize their DX11 driver for this game since what's the point of wasting their financial resources on 2 different APIs?! It's like would you run the game in DX if OpenGL produced better performance with no adverse side-effects? No, you wouldn't. It should be a matter of time before AMD releases a driver that enables CF support under Mantle.

I facepalm so hard when someone uses price to try to discuss the technical performance of something like mantle. Seriously, price is a function of supply and demand (and company internal policy and profit margins and the like) and HAS NO BEARING ON THE TECHNICAL ASPECTS OF MANTLE. Sure a company can charge more with more features but the 980 to 290X perf/price argument is heavily in favour of the 290X, mantle or no mantle. Nvidia's cards have always been more expensive, suddenly this means that DX is doomed and mantle reigns supreme?

Also people seem to be thinking that there is some kind of "CPU optimization for the driver for a specific game". This 100% is not completely true. It is very possible to streamline drivers for DX 11 in general (which is what Nvidia did) with further benefits from game specific (really engine specific) optimizations.

I like mantle. It seems to work and brings nice real world benefits. However, I cannot help but think that AMD would be capable of so much more with mantle if they were not under such budget constraints.