COD:AW Benchmarks

[guskline]

It is running well on my rig below. Was a blast playing it on my new BenQ BL3200PT:biggrin:

 

[desprado]

Well here is the real prof why Nvidia has much better Draw calls with their drivers compare to AMD and rest the video will explain.

https://www.youtube.com/watch?v=lQzLU4HWw2U

I am not trying to flame but i am just tying to post facts and thats all.

 

[SPBHM]

I've been using an i3 and AMD VGA for years, maybe that's why I always noticed a lot of CPU bottleneck and felt the need to recommend faster CPUs than what I have... it's quite a shame for AMD, when their VGAs offer better perf/$ only with higher end CPUs...

but this is something worth investigating more, can they reproduce this behavior in more games? what about different CPUs, could this be specific to the i3 (dual core with HT) compared to let's say a sandy bridge i5 at 3ghz...

 

[desprado]

I've been using an i3 and AMD VGA for years, maybe that's why I always noticed a lot of CPU bottleneck and felt the need to recommend faster CPUs than what I have... it's quite a shame for AMD, when their VGAs offer better perf/$ only with higher end CPUs...

but this is something worth investigating more, can they reproduce this behavior in more games? what about different CPUs, could this be specific to the i3 (dual core with HT) compared to let's say a sandy bridge i5 at 3ghz...

Well Gamegpu.ru has shown alot of example using GTX 780 TI vs R9 290X processor load min and max but some of here say that GTX 780 TI is faster and nothing at all but i want know their comment on this.

 

[RaulF]

Not sure why you felt the need to bring this up again. There's a thread about it already.

http://forums.anandtech.com/showthread.php?t=2406682

 

[desprado]

Not sure why you felt the need to bring this up again. There's a thread about it already.

http://forums.anandtech.com/showthread.php?t=2406682

This is different.I am only taking about draw calls and how Nvidia perform better on low end CPU on Dx that is all.

Some people said that GTX 780 Ti is faster and that is why it has higher CPU usage which was totally BS.

 

[RaulF]

No is the same. Is about COD AW and how Nvidia has much better DX drivers than AMD. It has been brought up on the thread I posted.

 

[desprado]

No is the same. Is about COD AW and how Nvidia has much better DX drivers than AMD. It has been brought up on the thread I posted.

Really than i can change the topic and post Civilization beyond earth Nvidia Dx11 vs AMD Dx11.U want that?

 

[Silverforce11]

I've been using an i3 and AMD VGA for years, maybe that's why I always noticed a lot of CPU bottleneck and felt the need to recommend faster CPUs than what I have... it's quite a shame for AMD, when their VGAs offer better perf/$ only with higher end CPUs...

but this is something worth investigating more, can they reproduce this behavior in more games? what about different CPUs, could this be specific to the i3 (dual core with HT) compared to let's say a sandy bridge i5 at 3ghz...

This has only been a recent situation, in particular after BF4 and Mantle came out, NV released an updated DX11 driver a few months after which dramatically boost CPU efficiency. It was their response to Mantle. But now it carries through in many DX11 games where NV enjoy lower CPU overhead.

I suspect some of the Total System power obtained from reviews in which there's a wide gap for 780ti vs R290X even-though measured via the GPU only, they are similar ~225W average in games is due to lower CPU overhead on NV DX11. The CPU works less to push the GPU, so there's extra power savings gained!

AMD definitely needs to push for better CPU overhead on DX11 as well, without going full-retard on Mantle-only optimizations for a few games that use it.

1. It will help systems with weaker CPUs (AMD CPUs especially!!) get better FPS (Min FPS).
2. Lower system power consumption.

ps. This is a VERY IMPORTANT topic and AMD needs to be 100% aware that they have a major problem on their hand with NV being so much more efficient on DX11.

 

[raghu78]

The last time I checked AMD R9 cards were matching or beating their Kepler counterparts in COD AW. btw other than computerbase which ran the benchmarks again with catalyst 14.11.1 beta which improved performance, the rest of the sites ran with 14.9.2 beta. So it would be even more favourable if the other sites ran with 14.11.1 beta. 1080p performance improved by 11% in 1080p and 2-3% in 1600p/4k in the computerbase test with catalyst 14.11.1 beta.

http://www.computerbase.de/2014-11/benchmarks-amd-treiber-cod-advanced-warfare/

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,6.html

http://www.sweclockers.com/artikel/19592-snabbtest-grafikprestanda-i-call-of-duty-advanced-warfare

http://www.benchmark.pl/testy_i_recenzje/call-of-duty-recenzja/strona/23024.html

http://www.pcgameshardware.de/Call-of-Duty-Advanced-Warfare-PC-258559/Specials/Test-Technik-1141322/

 

[desprado]

The last time I checked AMD R9 cards were matching or beating their Kepler counterparts in COD AW. btw other than computerbase which ran the benchmarks again with catalyst 14.11.1 beta which improved performance, the rest of the sites ran with 14.9.2 beta. So it would be even more favourable if the other sites ran with 14.11.1 beta

http://www.computerbase.de/2014-11/benchmarks-amd-treiber-cod-advanced-warfare/

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,6.html

http://www.sweclockers.com/artikel/19592-snabbtest-grafikprestanda-i-call-of-duty-advanced-warfare

http://www.benchmark.pl/testy_i_recenzje/call-of-duty-recenzja/strona/23024.html

http://www.pcgameshardware.de/Call-of-Duty-Advanced-Warfare-PC-258559/Specials/Test-Technik-1141322/

Yes R9 series is great but what i am really focusing in this topic how Nvidia cards perform with low end CPUs compare to AMD.

Most of those benchmark are taking from start of the game and game starts inside building meaning less draw calls.

 

[RaulF]

The point is that there is already a thread about COD Ghost and it has been talked about Nvidia performing better than AMD.

You made the call to make another to talk about it again.

But yet I am the troll.

 

[desprado]

The point is that there is already a thread about COD Ghost and it has been talked about Nvidia performing better than AMD.

You made the call to make another to talk about it again.

But yet I am the troll.

No i can show u benchmarks where even in BF4 Nvidia performances much better with low end cpus compare to AMD on DX11 and if u want that i can post for u and others game too.

 

[RaulF]

I see that you cannot comprehend what I am trying to type.

You are right and I'm wrong, good job you won.

Goodbye.

 

[KeithTalent]

Threads merged.

Anandtech Administrator
KeithTalent

 

[psolord]

I've been using an i3 and AMD VGA for years, maybe that's why I always noticed a lot of CPU bottleneck and felt the need to recommend faster CPUs than what I have... it's quite a shame for AMD, when their VGAs offer better perf/$ only with higher end CPUs...

but this is something worth investigating more, can they reproduce this behavior in more games? what about different CPUs, could this be specific to the i3 (dual core with HT) compared to let's say a sandy bridge i5 at 3ghz...

The problem here is not just that AMD gpus need higher end cpus. The problem is that whenever a cpu limit comes into play, AMD will yield faster than Nvidia.

For example, when recording the video game you play, a cpu that is more relaxed, will produce better recordings.

Also when gaming in stereoscopic 3D (ie Tridef) the cpu is hammered as well. Quite possibly with Nvidia's cards, things would be much much better.

Both the above are two things that I do often and I have seen my cpu limits first hand, with my 7950s.

This is the saddest thing I have read regarding AMD in recent years. They need to fix this yesterday.

 

[gamervivek]

Yeah this came up with those pclab.pl(?) benchmarks, but didn't realize that it would be this bad. r9 280 worse than 750ti is pretty shocking.

 

[desprado]

Yeah this came up with those pclab.pl(?) benchmarks, but didn't realize that it would be this bad. r9 280 worse than 750ti is pretty shocking.

on a slower cpu and while gameplay not 30 to 40 secs benchmark inside the building.

 

[Flapdrol1337]

Yeah this came up with those pclab.pl(?) benchmarks, but didn't realize that it would be this bad. r9 280 worse than 750ti is pretty shocking.

If it's the cpu holding back performance it's not shocking at all, any amd gpu would perform worse than the 750Ti in that case. With a badass cpu the 280 will still easily outperform the 750Ti.

 

[KaRLiToS]

This game is running like crap on my system. With resolution of 8040x1440 (triple monitors) and video settings all set to max, I get around 140fps but for some reason I get some big frame drops or kind of freeze during gameplay. It is making the game unplayable for me.

(I use widescreen fixer to make it work on 3 monitors because it is not supported out of the box)

And I tried 3840x2160 (4k) and it is working fine though with around 185fps.

But I still get some frame dips or some kind of freeze once in a while.

(Cards are Quad R9 290x running stock 1000/1250)

 

[psolord]

Hello, can someone explain why I see pagefile activity although I have disabled it?

MSI AB reports both RAM and Pagefile being used, although pagefile is disabled!

I think that if the game was not using pagefile at all, it would run faster in general.

 

[Ryanrenesis]

Hello, can someone explain why I see pagefile activity although I have disabled it?

MSI AB reports both RAM and Pagefile being used, although pagefile is disabled!

I think that if the game was not using pagefile at all, it would run faster in general.

"Pagefile" on MSI AB is not Windows pagefile but rather total RAM usage.

 

[psolord]

"Pagefile" on MSI AB is not Windows pagefile but rather total RAM usage.

Oh I see.

Still it has separate values for ram usage and pagefile.

My pagefile was indeed disabled though. I Alt+tabbed and there was no pagefile created on any of the drives.

 

[Ryanrenesis]

Oh I see.

Still it has separate values for ram usage and pagefile.

My pagefile was indeed disabled though. I Alt+tabbed and there was no pagefile created on any of the drives.

Yeah, creator of MSI Afterburner, Unwinder, has made some statements about it. Pagefile on MSI Afterburner does not reflect windows pagefile usage but RAM + something I forgot.

EDIT: My god, this game is using 10.8Gb of my RAM, 19.05Gb of Pagefile usage and somehow more vRAM than I have in my R9 290 Crossfire, 7.821Gb... errr WHAT??

It's running perfectly smooth 120 FPS everything max'd out supersampling 2x and 2x SMAA. I cannot ask for better performance.

 

[skipsneeky2]

Giving this game a try on the Steam free weekend,fun for the most part but damn this is one sloppy freaking port.Expected sure but why a game that looks so horrible tanks my 2gb 770 is beyond me.Another sloppy port that could look better but of course is gimped by the 2gb.Had to drop all kinds of settings to get the stuttering to disappear.

I know i checked the GameGpu review,but this 770 is matched with a 768p screen.I am not going for the 4k experience here with dual 980s,i just wonder where the hell all the vram goes too at this low of a resolution.

Someone be awesome enough to grab a picture of Captain Jack Sparrow from the Pirate movies and caption in"Why is the vram always gone" cause i pretty much can't see pass that image in my head right now.