OFFICIAL: BF4 Mantle Cat 14.1 Performance and Feedback Thread

B

Brent_Justice

Moderator
Joined
Apr 17, 2000
Messages
17,755
I'm creating this thread so that we have one thread dedicated to BF4 / Mantle / Cat 14.1 Performance and Feedback instead of everyone creating a million threads to discuss it.

We have our own preview article in the works.

Mantle is out in BF4 as of the latest patch.

Cat 14.1 is out which enables Mantle in BF4, you do have to select the Mantle API from within BF4 graphics options.

http://support.amd.com/en-us/kb-articles/Pages/latest-catalyst-windows-beta.aspx

Before Installing Cat 14.1 I suggest properly removing your previous AMD driver.

Make sure you properly removed all of your previous AMD driver, registry, driver cache INF storage area, everything.

This utility will do it: https://forums.geforce.com/default/...ver-uninstaller-ddu-v11-1-released-01-23-14-/

http://www.wagnardmobile.com/DDU/

--------------------------------------------------------------

AMD has also released more info:

Blog: “AMD Catalyst™ 14.1, a must-have driver for gamers!” - http://community.amd.com/community/...md-catalyst-141-a-must-have-driver-for-gamers

Known Issues: http://support.amd.com/en-us/kb-articles/Pages/Mantle-Known-Issues.aspx

FAQ: “Mantle Graphics API Frequently Asked Questions”: www.amd.com/mantlefaq

------------------------------------------------------------

You can discuss driver specific issues related to 14.1 in this thread

http://hardforum.com/showthread.php?t=1804584

Let's leave this one for Mantle Performance comparisons, benchmarks, and feedback.

--------------------------------------------------------------

Post your Performance differences.

Post your feedback on Mantle and BF4.

Have fun gaming!

-Brent
 
Last edited:
The difference is enormous in BF4 for me, performance is close to doubled (and I'm STILL CPU limited according to the graph!). I'm being blown the fuck away. I played an obliteration round on Shanghai and it was like 60+ mostly. It was nearly unplayable before, I'd just leave when it came on because it ran like shit. A round on Locker was more often around 100~... and C didn't put a massive dent in my framerate whenever I got near it.

Q6600 @ 3ghz, 7870

Shit is bonkers for me.

edit: Also Star Swarm performance:

me said:
Performance difference is looking pretty fucking huge on my system...

FPS low point was about 16-17~ ish on Mantle.
FPS low point was about 3-4~ ish on DX

The demo doesn't seem to play the exact same every time, but the difference was fucking blatant. Mantle was MUCH smoother throughout.

Q6600 @ 3.0
1100 / something 7870

Mantle:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 11742

Average FPS: 32.61
Average Unit Count: 3900
Maximum Unit Count: 5386
Average Batches/MS: 576.65
Maximum Batches/MS: 1339.42
Average Batch Count: 20801
Maximum Batch Count: 118615
===========================================================

DX inc
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 5859

Average FPS: 16.27
Average Unit Count: 4176
Maximum Unit Count: 5688
Average Batches/MS: 245.99
Maximum Batches/MS: 577.00
Average Batch Count: 20272
Maximum Batch Count: 136316
===========================================================
Click to expand...
 
No crossfire here. I see good FPS gains but I get frame drops and stutters in large maps. No so much in Domination maps.
I also have an issue with the mouse cursor where it wont be accurate when I click on something. I have to lower my mouse down below the button I want to press in order for it to be activated.

I noticed the GPU Memory Usage has doubled. Almost to 3GB from 1.5 under DX11

3570k @ 4.5
Sapphire 7950 @ 1150/1400
 
I was getting a kind of stutter before too. Lowered texture quality from Ultra to High -> seems to have fixed it.

And yeah, the UI is kind of borked for some reason. If you alt-enter to window'd mode it's accurate.
 
12-13% boost on a 2600k and single 7950 using a custom mix of settings low/high/ultra using Mantle

Having sudden fps drops down to single digits every ten minutes or so is the only downside other than DXtory and FRAPs not working with it.
 
Great performance for me.
Only played 1 or 2 hours all seemed great and no glitches and no stuttering

Rig I was playing on.

3770K 4600MHZ
R290+680GTX
1920x1080p Ultra 4xMSAA max

SP-25 FPS Increase
MP- 3 different maps-around 10-20 FPS Increase
 

socK, what resolution and quality setting with BF4, 1080p ultra?
 
Last edited:
Interesting results here. Running 2600K w/R9 290 in X-Fire.

In Star Swarm with X-fire:

DX
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 8575

Average FPS: 23.82
Average Unit Count: 4262
Maximum Unit Count: 5539
Average Batches/MS: 482.70
Maximum Batches/MS: 965.30
Average Batch Count: 23340
Maximum Batch Count: 154919
===========================================================

Mantle
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 20001

Average FPS: 55.56
Average Unit Count: 4366
Maximum Unit Count: 5597
Average Batches/MS: 1051.40
Maximum Batches/MS: 2651.54
Average Batch Count: 22176
Maximum Batch Count: 97733
===========================================================

In Star Swarm X-Fire disabled:

DX:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 14117

Average FPS: 39.21
Average Unit Count: 4371
Maximum Unit Count: 5448
Average Batches/MS: 643.54
Maximum Batches/MS: 1476.72
Average Batch Count: 17920
Maximum Batch Count: 104127
===========================================================

Mantle:
Star Swarm crash on start...won't run.

Enabled X-Fire and it works again.

*Edit*

Downgraded to 13.12 and tried Swarm in DX

13.12 DX X-fire on:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 9205

Average FPS: 25.57
Average Unit Count: 3983
Maximum Unit Count: 5602
Average Batches/MS: 432.23
Maximum Batches/MS: 838.92
Average Batch Count: 18266
Maximum Batch Count: 113385
===========================================================

13.12 DX X-Fire Disabled:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 15098

Average FPS: 41.94
Average Unit Count: 4383
Maximum Unit Count: 5573
Average Batches/MS: 651.79
Maximum Batches/MS: 1488.26
Average Batch Count: 17231
Maximum Batch Count: 97311
===========================================================
 
Last edited:
Here's a quick chart I threw together from my frame time log in BF4 through a 64-player multiplayer match on Altai Range. Switching from DirectX to Mantle I saw higher overall FPS and higher minimum FPS (never dropped below 60), but I haven't yet recorded FPS.

"Threads" on Render.DrawScreenInfo went from 2 to 3 with Mantle enabled

Enabling Mantle meant I had to disable Intel integrated graphics in Device Manager or BF4 would freeze/crash when trying to enter the Options menu.

I also noticed periodic stuttering with Mantle when playing. The large spikes on the Mantle part of the chart actually go up quite high, but I kept the max value at 70 to match with the DirectX chart. While not deal-breaking, they were certainly super annoying.

i5 3570k @ 4.2GHz, R9 290 stock clock, 1600x900 Ultra preset.

Untitled_zps639aecab.png
 
Last edited:
Here's my post from the previous thread:

Just did a multiplayer test run on a 64-player Siege of Shanghai map. Specs: 4770k at 4.3 GHz and a 290X.

I play at 1080p, 135% resolution scale, Textures Ultra, everything else Low/Off. Did my best to do the same stuff each match. Either way, results are pretty significant after an 8 minute benchmark run:

DX11: 100 FPS
Mantle: 142 FPS
Click to expand...
 
more memory being used is not at all a bad thing if it has the performance to back it up, what are all your temps like and have you noticed any major issues with Cat 14.1 and/or those scrypt mining if you are any performance boosts or losses with 14.1?

I want to update to these I would just hate to "suffer" :p
 
Subbed, if anyone know how to solve memory leak issues, please let me know (pm) so i can test it :eek:

Game is crashing!
 
socK said:
edit: Also Star Swarm performance:
Click to expand...
On a related note, I don't know how they're going to get any kind of multiplayer component working in Oxide without a deterministic simulation. They certainly can't send state packets for tens of thousands of objects even a few times per second.
 
wonderfield said:
On a related note, I don't know how they're going to get any kind of multiplayer component working in Oxide without a deterministic simulation. They certainly can't send state packets for tens of thousands of objects even a few times per second.
Click to expand...

Well, I guess they have it figured out

Q. Can I make a multiplayer game with Nitrous?
A. Yes, Nitrous supports Steam out of the box. Since Nitrous is very configurable, the options for multiplayer are wide open. Due to how gameplay code interfaces with the engine it is extremely easy to replicate state data across machines.

Q. Can I make a massively multiplayer game with Nitrous?
A. Yes. Nitrous is designed to handle very heavy workloads like you would see in a massively multiplayer game. From high unit counts in gathering places to lush and full environments, we would love to help bring your world to life.
 
I was getting a drop spike once in awhile under mantle on 32bit windows 8.1 .. so I switched to 64bit and the spikes was gone. game runs great and looks great as i'm on 1920 x 1080 Ultra with 120% scaling and 4xMSAA and post processing is set to high.

I see maybe 6 to 8 fps more under Mantle ,, with i7-930 @3.4 Ghz and R9-290 stock
 
Here is one of my starswarm results: Went with the "follow" scenario and the extreme setting with the mantle render pathway.

w1t4.jpg


== Hardware Configuration =================================
GPU: AMD Radeon R9 200 Series
CPU: AuthenticAMD
AMD FX(tm)-8350 Eight-Core Processor
Physical Cores: 4
Logical Cores: 8
Physical Memory: 8532619264
Allocatable Memory: 8796092891136
===========================================================


== Configuration ==========================================
API: Mantle
Scenario: ScenarioFollow.csv
User Input: Disabled
Resolution: 1920x1080
Fullscreen: True
GameCore Update: 16.6 ms
Bloom Quality: High
PointLight Quality: High
ToneCurve Quality: High
Glare Overdraw: 16
Shading Samples: 64
Shade Quality: Mid
Deferred Contexts: Disabled
Temporal AA Duration: 16
Temporal AA Time Slice: 2
Detailed Frame Info: Off
===========================================================


== Results ================================================
Test Duration: 360 Seconds
Total Frames: 18705

Average FPS: 51.96
Average Unit Count: 4514
Maximum Unit Count: 5735
Average Batches/MS: 801.93
Maximum Batches/MS: 1768.72
Average Batch Count: 17953
Maximum Batch Count: 113230
===========================================================


Still saw some impressive slowdowns once the counts got high and the entire fleet swarm was in view with all the lights, but nothing compared to dx11, that was a slideshow.
 
same deal, massive slowdowns with motion blur and huge units. nice 20fps or so on min.

d3d was a sllide show....7fps at points...
 
Im seeing big gains in bf4 multi player. Seige of shanghai never dropped below 60. Most of the time it was between 95-120 fps. This is on a 3770k and 290x. Every map ran smooth with the occasional stutter. Im super happy with the results!
 
Here are my starswarm results (Scenario: Follow, Settings: Extreme, Timed Run) with a Core i7 920 c0 at 3.33GHz and a R9-290X at stock clocks.

== Hardware Configuration =================================
GPU: AMD Radeon R9 200 Series
CPU: GenuineIntel
Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
Physical Cores: 4
Logical Cores: 8
Physical Memory: 12874534912
Allocatable Memory: 8796092891136
===========================================================

AMD 13.11 D3D:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 10068

Average FPS: 27.95
Average Unit Count: 4113
Maximum Unit Count: 5476
Average Batches/MS: 532.09
Maximum Batches/MS: 1051.12
Average Batch Count: 21469
Maximum Batch Count: 84505
===========================================================

AMD 14.1 D3D:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 10853

Average FPS: 30.15
Average Unit Count: 4050
Maximum Unit Count: 5554
Average Batches/MS: 537.13
Maximum Batches/MS: 1035.93
Average Batch Count: 19648
Maximum Batch Count: 126819
===========================================================

AMD 14.1 Mantle:
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 20010

Average FPS: 55.58
Average Unit Count: 4204
Maximum Unit Count: 5488
Average Batches/MS: 844.25
Maximum Batches/MS: 1993.91
Average Batch Count: 17686
Maximum Batch Count: 80321
===========================================================
 
This is working so well for me. Pretty much played all night.

Limited gains they said, not optimized for 7 series hardware they said. :p
 
But if that's the case, why isn't Mantle running significantly faster in non-CPU bottlenecked scenarios too?
 
To install this driver, I first uninstalled the 13.12whql with the official AMD uninstaller, via ctrl panel. Then I rebooted to safe mode and ran Driver Fusion and then immediately followed it with Display Driver Uninstaller, which initiates an auto-reboot. I then installed the AMD chipset drivers appropriate for my mobo (because I noticed the 14.1 betas don't include the USB filter driver etc), rebooted, and then I installed the 14.1 betas.

My system is the following:

HD 7870 overclocked to 1150 core and 1325 memory
Phenom II X6 overclocked to 3.6ghz
8gigs of ram
Soundblaster X-Fi

Windows 7 64-bit, running BF4 in 64-bit mode.

with these settings (these are my daily play settings. I did not pic these settings special for these tests. This is how I have been playing BF4, for two months):

(I forgot to take a "Mantle" screenshot of settings. Pretend it says Mantle instead of Direct X 11)
ScreenshotWin32-0002.png


I get these image quality differences on the test server and I also noticed similar issues with higher brightness and fogging in Operation Locker and Siege of Shanghai ( the only two maps I have played since Mantle installation):

Direct X 11
ScreenshotWin32-0003.png


Mantle
ScreenshotWin32-0004.png


DirectX 11
ScreenshotWin32-0008.png


Mantle
ScreenshotWin32-0006.png


DirectX 11
ScreenshotWin32-0010.png


Mantle
ScreenshotWin32-00011.png


*these images were taken with BF4's own internal screenshot command. Here are the original, untouched PNG files:
https://www.dropbox.com/s/ghku2sj3kyxbb9c/bf4MantleVSdx11testserver.zip


Additionally, Mantle has very noticeable hitching. The more players in the map, the worse it is. On the Test server, it's mild and irregular. On 64 player Siege of Shanghai, it hitches every few seconds, quite noticeable, and occasionally it will hitch super hard and drop several frames. Kinda reminds me of playing the BF4 PC beta...

The DX11 codepath in 14.1 seems to be ok. Though I'll have to play with it more. with 13.12's would give me lighting issues sometimes, after looking through scopes. particularly infrared scopes. After returning to normal/hip view, the lighting would appear very basic for a second and then somewhat quickly bleed back into high quality lighting.

I have frame data from perfoverlay that I will be uploading later. as well as CPU cores and GPU utilization figures from AMD System Monitor.

*Here is the frame data:

I did frame data only, on the test server for the 13.12, 14.1 DX11, and 14.1 Mantle. 13.12 had just been re-installed two days ago, using both driver cleaners mentioned near the beginning of this post. The data is based on a lap around the test server where I shoot a bunch of stuff with an LMG, C4 several things, fly a chopper and shoot some stuff, crash the chopper, ride an ATV, etc. I took care to shoot all the same things, C4 the same things with the same amounts of C4, crash chopper in same spot, etc. As little variation as I could muster.
Here is a quick H.264 encode of one such lap, that I did after all these tests. To capture, I used the MJPEG codec in MSI Afterburner, using BF4's 32-bit mode, via DX11 codepath (Afterburner was never running during any of the other tests). The encode is not meant to be pristine quality. Just quick and dirty, to see what the lap looks like. You can watch it streaming, or download the original file, which looks better:
https://www.dropbox.com/s/wgd3gr8fdlmn429/bf4testserverrunsamplea.mp4

I did both frame data and CPU core/GPU utilization at the same time, on Siege of Shanghai 64 player. All for 13.12whql, 14.1 DX11, and 14.1 mantle (I used a 24/7 Siege of Shanghai only server with 60's average ping. That way I could run the tests all on the same server and map). Each test was roughly 6 minutes. ( set a timer for 6 minutes, but some extra time was added to start and stop both perfoverlay and AMD System Monitor.)

All frame data files:
https://www.dropbox.com/s/ou43z6ywg0rl0co/Battlefield4 frame data.zip

All CPU/GPU utilization files:
https://www.dropbox.com/s/e8bi6w39khsymhh/Battlefield4cpuGPUutilization.zip

--------------------------------------------

Here are the average frame rates for the test server lap:

14.1 Mantle: 108
14.1 DX11: 103
13.12 DX11: 104

You can see here, the Test server probably doesn't stress CPU performance. As such, the average performance is basically the same. There are points at which Mantle performs better in the Test Server. But overall it is pretty much the same.

-----------------------------------------------------------

Average frame rates for 64 player Siege of Shanghai, during 6 minutes of play:

14.1 Mantle: 85
14.1 DX11: 58
13.12 DX11: 54


Here is where we get HUGE gains. That speaks for itself. Even despite the fact that the Mantle run was hitching every few seconds, with occasional hard stops/several frame drops, it still turns in a much higher average number.

-----------------------------------------------------

**UPDATE

looks like the stuttering is very likely due to Mantle using quite a bit more VRAM space, for the same settings:

EnhancedInterrogator said:
It really looks like Mantle (on the 7000's) is using more VRAM, or using it less efficiently, given how it responds to ultra textures
Click to expand...

CONFIRMED to be an issue (at least for some of us...)

using the same system and the exact settings I I have been using all along:


here is vram usage compared between DX11 and Mantle, on my 7870:

*these shots were taken soon after loading into the test server.

DX11 screen 1
https://www.dropbox.com/s/352up4h89lvcq90/dx11vmemtest1.jpg

Mantle screen 1
https://www.dropbox.com/s/bmarkleomxbinkp/Mantlevmemtest1.jpg

DX11 screen 2
https://www.dropbox.com/s/oo2xpxwo1ed6von/DX11vmemtest2.jpg

Mantle screen 2
https://www.dropbox.com/s/3u3h6e7oxkm309b/Mantlevmemtest2.jpg

*also captured in these screenshots, Mantle makes my card run way hotter.

**I took a few more screens. This time after I'd been in the server for several minutes. I loaded in, ran my usual lap (which takes a few minutes) and then took a couple of shots. Vram usage is even higher. DX11 temps top out at 58. Mantle pushes it even higher still...

DX11 ground shot
https://www.dropbox.com/s/k43noxu0ns0azlb/1_DX11vmemtestafterseveralminutes.jpg

Mantle ground shot
https://www.dropbox.com/s/jpp0vnxu1icueg6/2_Mantlevmemtestafterseveralminutes.jpg

DX11 3rd person chopper shot
https://www.dropbox.com/s/ipifdhy5jv1a43t/3_DX11vmemtestsveral minutes chopper.jpg

Mantle 3rd person chopper shot
https://www.dropbox.com/s/yf2pgpega6o7c9y/4_Mantlevmemtestafterseveralminuteschopper.jpg

**regarding the fan speed in those shots:

I have a HIS card that had the exhaust style turbine blower setup. In my experience, those never push enough air. The card used to hit 68C on a custom fan curve, that pretty much cranked it once it hit 55C. I had also applied Arctic Silver Ceramique.
I tore the exhaust shroud and turbine fan off and twist tied two 80mm case fans to the top of the heatsink. The fans are plugged into the mobo, so I just turned the videocard's fan speed setting as low as it would go. Better temps, less noise.

I did a similar mod to an Nvidia 6800 I had 10 years ago.


GPU-Z logs tracked during two test laps of the server, mainly to show/compare VRAM usage over time. but it tracks several other stats, as well:

one lap with 14.1 DX11 and one lap with 14.1 using the Mantle option:

https://www.dropbox.com/s/zdtencorj2vjt3q/14_1VRAMtestserverlapsGPUz.zip


Here is a low quality video simply showing what the laps look like. You can watch it streaming, or download the original file which looks better:
https://www.dropbox.com/s/wgd3gr8fdlmn429/bf4testserverrunsamplea.mp4

The lap in that video was taken right after I ran the laps that logged the VRAM data. All three laps are very similar. Close enough, that you could use the raw data to see exactly what's happening with the GPU VS. the on screen action.

-------------------------------------------------------

*slightly off topic: it would be nice if they would put the super ammo box back into the test server. I haven't been able to successfully reload any of the rocket launchers or the grenade launcher that you pickup off the ground. The shotguns reload off an Support pack, but it was nice to be able to load in 100 bullets/rocket/grenades/etc and just go for it. Makes doing test runs a lot easier.
 
Last edited:
I was finally able to change my crossfire setup to use directx instead of mantle, and mantle would crash my pc as soon as I loaded bf.
also, before this new driver I was using resolution scaling with no problem...this driver seems to wreak havoc when I use it. so, I set it back to 100 and then added 16xSSAA in CCC and the game work fine, as pre-14.1 driver. however, mantle is broken.
 
went from 16.5fps avg with swarm on attract/extreme to 46fps avg with mantle.

BF4 also runs like fuckin butter now with everything maxed. I dont notice any drops in FPS. Although i didnt benchmark prior, i could definitely notice the difference.
 
Haven't noticed any visual differences, but then again I'm not going through it with a fine toothed comb. The game is considerably smoother though, even in areas that used to be a drag even on my system. Minimum fps saw a huge jump, as did average fps. Even despite their claim that the HD 7000 people wouldn't see that big a jump, turned out to be false, its very considerable.

The people with crossfire will have an issue because with Mantle, crossfire support is needed is handled by the game it seems and not the driver anymore.
 
gordon151 said:
Haven't noticed any visual differences, but then again I'm not going through it with a fine toothed comb. The game is considerably smoother though, even in areas that used to be a drag even on my system. Minimum fps saw a huge jump, as did average fps. Even despite their claim that the HD 7000 people wouldn't see that big a jump, turned out to be false, its very considerable.

The people with crossfire will have an issue because with Mantle, crossfire support is needed is handled by the game it seems and not the driver anymore.
Click to expand...

That's what I was wondering. Can anyone confirm that multi-gpu is working in BF4 under Mantle with crossfire disabled in the drivers?

From what I've read, it seems like it does work for some gpus. Probably only works with GCN 2 gpus if I'd hazard an uniformed guess.
 
Pretty awesome. I'm impressed by both the BF4 patch and the drivers released last night. I played a solid four hours(with my awesome patient girlfriend reading on her kindle... and eventually falling asleep! more wine for me!) last night. In the four hours I had one crash(which was probably caused from me alt+tabbing out a bunch).

With vsync enabled in BF4, ultra settings, and 1920x1080... I can honest to god say no matter what map, what the weather conditions were, or even having 64players... I never saw it change from 60fps.

With vsync disabled the lowest I saw it drop was in the low 80s... with an average of 110-115.

Star Swarm results:

===========================================================
Oxide Games
Star Swarm Stress Test - ©2013
C:\Users\Joel\Star Swarm\Output_14_02_02_0149.txt
Version 1.00
02/02/2014 01:49
===========================================================

== Hardware Configuration =================================
GPU: AMD Radeon R9 200 Series
CPU: AuthenticAMD
AMD Phenom(tm) II X6 1100T Processor
Physical Cores: 6
Logical Cores: 6
Physical Memory: 17139200000
Allocatable Memory: 140737488224256
===========================================================


== Configuration ==========================================
API: Mantle
Scenario: ScenarioFollow.csv
User Input: Disabled
Resolution: 1920x1080
Fullscreen: True
GameCore Update: 16.6 ms
Bloom Quality: High
PointLight Quality: High
ToneCurve Quality: High
Glare Overdraw: 16
Shading Samples: 64
Shade Quality: Mid
Deferred Contexts: Disabled
Temporal AA Duration: 16
Temporal AA Time Slice: 2
Detailed Frame Info: Off
===========================================================


== Results ================================================
Test Duration: 360 Seconds
Total Frames: 16639

Average FPS: 46.22
Average Unit Count: 4349
Maximum Unit Count: 5424
Average Batches/MS: 801.49
Maximum Batches/MS: 1843.61
Average Batch Count: 21837
Maximum Batch Count: 114407
===========================================================


No overclocks... other than cpu being at 4.1ghz
 
Well, all the Mantel threads are being closed and consolidated and there isn't a Star Swarm specific thread so it seems fair game.

As for BF4, seeing slower performance at 1080p ultra but might be seeing ~15% higher at low but have to do further tests.

Do see about doubling in performance in Star Swarm:

DirectX
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 5585

Average FPS: 15.51
Average Unit Count: 3773
Maximum Unit Count: 5460
Average Batches/MS: 300.56
Maximum Batches/MS: 521.96
Average Batch Count: 22476
Maximum Batch Count: 160741
===========================================================

Mantel
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 13039

Average FPS: 36.22
Average Unit Count: 4110
Maximum Unit Count: 5519
Average Batches/MS: 542.26
Maximum Batches/MS: 1271.89
Average Batch Count: 17479
Maximum Batch Count: 115098
===========================================================
 
Why are the official reviews from big sites not out yet? why no word on the destruction 14.1 drivers cause with crossfire? why do some people say their crossfire works excellent with mantle? 14.1 destroyed my directx perf and bf4 is unplayable under mantle API. I am very angry.
 
Intel i7 940 @3.53ghz
6gb DDR3 1600mhz
Sapphire 290 Tri-x

Star Swarm Scenario "Follow.csv" Extreme:

DirectX
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 10828

Average FPS: 30.07
Average Unit Count: 4126
Maximum Unit Count: 5522
Average Batches/MS: 487.17
Maximum Batches/MS: 1073.33
Average Batch Count: 18078
Maximum Batch Count: 139151

Mantle
== Results ================================================
Test Duration: 360 Seconds
Total Frames: 21743

Average FPS: 60.39
Average Unit Count: 4403
Maximum Unit Count: 5465
Average Batches/MS: 849.98
Maximum Batches/MS: 2155.27
Average Batch Count: 16206
Maximum Batch Count: 164682
 
You must log in or register to reply here.
Back
Top