Catalyst 14.1 (MANTLE) Drivers are Here! Post yer Results!

[virus464]

germans tested 2, and some intresting slider picture are there

 

http://www.pcgameshardware.de/AMD-Radeon-Grafikkarte-255597/Specials/AMD-Mantle-Test-Battlefield-4-1107754/?utm_source=facebook.com&utm_medium=socialPCGH&utm_campaign=computecsocial&utm_content=AMD-Radeon-Grafikkarte-255597SpecialsAMD-Mantle-Test-Battlefield-4-1107754

[Fenzee]

2500k at 4.6Ghz, 7950 1100/1500

 

 

Running:

 

Follow, Extreme Settings

 

 

DX11

 

== Hardware Configuration =================================

GPU: AMD Radeon HD 7900 Series

CPU: GenuineIntel

      Intel® Core i5-2500K CPU @ 3.30GHz

Physical Cores: 4

Logical Cores: 4

Physical Memory: 17121746944

Allocatable Memory: 8796092891136

 

== Results ================================================

Test Duration: 360 Seconds

Total Frames: 15114

 

Average FPS: 41.98

Average Unit Count: 4252

Maximum Unit Count: 5569

Average Batches/MS: 573.06

Maximum Batches/MS: 1491.01

Average Batch Count: 15440

Maximum Batch Count: 66462

 

 

Mantle

 

== Hardware Configuration =================================

GPU: AMD Radeon HD 7900 Series

CPU: GenuineIntel

      Intel® Core i5-2500K CPU @ 3.30GHz

Physical Cores: 4

Logical Cores: 4

Physical Memory: 17121746944

Allocatable Memory: 8796092891136

 

== Results ================================================

Test Duration: 360 Seconds

Total Frames: 19135

 

Average FPS: 53.15

Average Unit Count: 4446

Maximum Unit Count: 5613

Average Batches/MS: 958.90

Maximum Batches/MS: 2920.56

Average Batch Count: 20973

Maximum Batch Count: 169395

 

 

About 27%~ increase in performance.

 

 

[pixeldensity]

So here my experience with mantle right now.

 

Specs: AMD Phenom X4 965 @ 3.4, XFX HD 7870, 8GB RAM

 

I was playing on Goumond Railway. Yes there was a increase in FPS, maybe 8 - 10 fps. However, there is a problem. During multiple sessions I notice that at intense points and even random points it would lag up to the point where it looked like it just froze. That is a major problem. Also what I expect from mantle wasn't just a fps increase I wanted a consistent fps, with no major drop in fps. I hope that once it is out of beta I can see my hopes come true. 

Actually it gave me 15 - 20 fps increase, my bad didn't realize you had to restart bf4 after switching. From a low 25 to 40 fps to  50 and above. However, I still hope they can hammer out the bugs that makes it freeze up every time in intense fire fights and other scenarios b/c freezing up in the middle of a large fire fight is a big problem in games especially in multiplayer and make the fps have more consistency without major fps drops.  

[nokturrduk84]

The things I wanted to say about Mantle are better spoken with this blog's article:

http://scalibq.wordpress.com/2014/02/02/independent-mantle-benchmarks-start-to-trickle-in/

 

I still think it won't have a future. It's even a pointless.

[Fetzie]

The things I wanted to say about Mantle are better spoken with this blog's article:

http://scalibq.wordpress.com/2014/02/02/independent-mantle-benchmarks-start-to-trickle-in/

 

I still think it won't have a future. It's even a pointless.

 

I disagree.

 

Firstly, all these BF4 benchmarks are from the GPU bound solo campaign. You see rather better numbers from the multi-player maps, which are more CPU bound. Mantle would be really great for an MMOG, because they are almost always dependent on having a strong CPU. RTS developers could finally leave the idea that an "army" has less than 20 units in it behind.

 

Indeed, it is in the interest of all computer gamers that we either get DX11 or OpenGl up to scratch (because they aren't) or Mantle gets adopted.

[Veltrix]

For my personal own right I have an Intel and Nvidia Grahpics card but I built my brothers with an AMD FX 6300 and HD 7850. I don't have any BF4 keys but I did get to use the benchmark StarSwarm. I ran it on my brothers all AMD rig. Here are the Results (this was on the Extreme Preset, Follow.csv, Time Run [6minutes])

                                                    D3D11(DX11) API                Mantle API                       Performance Gain

Total time                            =             360s                                   360s                                         x

Total number of Frames     =           ~7000fps                         ~13000fps                                  86%+

Average FPS                      =            19.96fps                            37.67fps                                  89%+

Lowest observed Fps         =               5fps                                  18fps                                    260%+    <- Lol but that's what I calculated

 

Keep in mind this is just a synthetic benchmark but that pretty impressive and watching through the entire benchmark I watched what the lowest fps I could observe. I wish I had BF4 to benchmark and to see a real world performance boost from the Mantle API. I hope also to see more games utilize mantle API and I get to do more benchmarking and more fun to see the performance gain. Well that's I have to show with the new Mantle API.

 

Full System Spec:

 

CPU:                         AMD FX 6300 Oc'd @ 4.00GHz

Cooler:                     Corsair A70 with x1 BitFenix Spectre pro PWM fan

Motherboard:            Asus M5A97 R2.0

Ram:                        Samsung 6Gb (2x2Gb + 2x1Gb) 1033Mhz RAM <- Reused RAM from an Acer computer Need to upgrade that for him soon
GPU:                        Sapphire AMD HD 7850 2GB GDDR5

OS:                          Windows 7 64-bit

PSU:                        Thermaltake TR2 500W 

Case:                      Zalman Z11 plus

HDD:                       1x WD 640GB Black drive / 1x WD 1TB green drive

 

[lhikary]

MANTLE: The Way The Hardware Is Meant To Be Used

Nvidia continue to bask in their PC gaming lead - a tiny 0.34% increase building on a more substantial 52.01% share

 

source:http://www.pcgamer.com/2014/02/03/steam-hardware-survey-shows-small-gains-for-windows-8-and-nvidia-through-january/?utm_source=fb&utm_medium=emp

 

[Sky Daddy]

CPU performance is the one major thing that sets Mantle aside from opengl/directx.

Planetside 2 should use mantle, just to see the performance difference of their 2,000 player game.

[1funnygame]

Lol when you have Mantle enabled the game uses 1 more Job thread of your CPU. (But still not all )

DX11.1:

http://postimg.org/image/xacqmyser/full/

Mantle:

http://s29.postimg.org/cz5heepg5/Screenshot_Win32_0002.png

whats with the 2 pixels less than 1080p?

[BroliviaWilde]

My experience/results are as follows:

I played for several hours of mutiplayer this past weekend in BF4 on multiple different maps switching between mantle and dx11.  In both situations I had settings on ultra with one step above minimum AA @ 1920 by 1080 with vsync off. I gotta tell you, screen tearing in this game is no where near as bad as it was in bf3 with either api.

 

EDIT: I gotta tell you, screen tearing in this game with either api is no where near as bad as it was in bf3.

 

Using mantle I noticed fps would fluctuate between about 65 - 72, using dx11 it was about 55 - 60, it wasn't always like this, but it was like this most of the time. In a few cases, they were the same, but never reversed. If I had never turned perfoverlay.drawfps, in all honesty I would have never noticed the perf increase, but it is nice to know it is there.

My setup runs at default settings, I'm lazy when it comes to OC and don't want to put the time into it:  2600k, 7970 16gb 2133 ram, 250gb SSD, rest of setup is typical stuff, blah blah blah, hence I don't have pics to back it up either.

Mantle is delivering more performance more often than not, so I know as the drivers are further optimized it will only get better.  But the biggest problem I've had is the game crashing at random times using either api.  Strongly thinking about going back to previous driver until the bugs get worked out a bit better so game won't crash.