Performance Benchmarks

[lotech]

Heres a test of the same machine in both OS's. Windows makes a huge difference in FPS on the same hardware.

Model: Mac Pro Late 2013
Graphic card: 2x AMD FirePro D700 6144 MB (total 12288mb)
Harddisk: Apple Flash storage 1Tb
Processor: 3,0 GHz 8-Core Intel Xeon E5
Memory: 32 GB 1867 MHz DDR3 ECC
Resolume 4.1.8

Output Setup: 1 Control monitor at 2560x1440, 1 Fullscreen output at 1920x1200

OS: Windows 8.1 Test with both CrossFire mode enabled (and forced) and disabled.
Benchmark for 1920x1080
Clean content: 22 layers
Noise content: 20 layers

OS: OS-X 10.9.2
Benchmark for 1920x1080
Clean content: 16 layers
Noise content: 10 layers

[He2neg]

System :

• Model: XMG P503 Schenker Notebook
  System: Windows 7 Pro (64bit)
  Processor: i7-4800MQ @2,70 Ghz
  Memory: 16 GB GB DDR
  GPU: Nvidia GTX 770M 3GB (Single GPU)
  Harddisk: Samsung 840 Pro 256GB

Resolume 4.1.11

SetUp:
- Laptop Monitor ( 1920 x 1080)
-Output Projektor ( 1920x1080p)

(Composition 640x480)
480p (clean) : did not testet xD
480p (noise) : 20 ( 60fps) / Gpu Load 39%

(Composition 1920x1080)
1080p (clean): 20 (40 fps) / GPU load 90% ( Gpu Z)
1080p (noise) : 20 (32 fps) / Gpuload : 96% (gpu Z)

[Somehow i fixed the Problem, i guessed it was something with the Intel optimus or the energy stuff]
Extra test :
1080p Noise / 20 Layers + colerize (timeline hue) + Delay RGB on max(0/0,5 / 1) ---> 30fps / 90-99% Gpu load

[cosmowe]

Notebook Model: DELL Alienware M18x R2
Operating System: Windows 7 Professional 64 Bit
GPU: AMD Radeon HD 7970 M, 2048 MB
Harddisk: Plextor PX-256M3P SSD Raid-0 , 512 GB
Processor: Intel i7 3820QM @ 2,7 GHz
Memory: 8GB RAM
Resolume: 4.1.8

Output Setup: 1x GUI at 1920 x 1080 60Hz, 1x Fullscreen Output at 1920x1080 60Hz

Benchmark 1920x1080
Minimal content: 20+ layers (20 @ 46 fps)
Noise content: 20 layers (20 @ 31 fps)

Benchmark 640x480
Minimal content: 20+
Noise content: 20+

Output Setup: 1x GUI at 1920 x 1080 60Hz, 2x Fullscreen Output at 1920x1080 60Hz

Benchmark 1920x1080
Minimal content: 20+ layers (20 @ 45 fps)
Noise content: 20 layers (20 @ 30 fps)

Benchmark 640x480
Minimal content: 20+
Noise content: 20+


Greetings
cosmowe

[jellobot]

Model: Alienware 18 (2014)

Operating System: Windows 7 64 bit
Graphic card: Dual NVIDIA GE Force GTX780
Harddisk: 80GB SSD
Processor: Intel Core i7 4900
Memory: 32 GB
Output Setup: Tested on 1 control monitor at 1920x1080 and two fullscreen outputs at 1920x1080


Benchmark for 1920x1080
Clean content: 20+
Noise content: 20+

Benchmark for 640x480
Clean content: 20+
Noise content: 20+

I'm really very surprised at the outcome ! Thank you for the benchmark tests Joris, I've always wondered about this! I've primarily been working 640x480 all this time but will definitely now consider going the full HD route!

**As noted by RenderRabbit Cosmowe, Resolume as of 4.1.8 does not yet take advantage of the dual graphics cards in this laptop, so this was based off of just one of the card's rendering...as you can see from either of our results the Alienwares seem to pair quite nicely with Resolume!

Cheers!

[Virus]

Model: Schenker XMG 702 Pro

Operating System: Windows 7 professional 64 bit on 7200 rpm HD
Graphic card: NVIDIA GE Force GTX675M
Harddisk: Samsung 256GB SSD
Processor: Intel Core i7 2670QM 2.2 GHz
Memory: 8 GB
Output Setup: 1920x1080 Monitor, 1920x1080 Screen


Benchmark for 1920x1080
Clean content: 20+ (39 fps)
Noise content: 16

Benchmark for 640x480
Clean content: 20+
Noise content: 20+

[Oaktown]

Model: MacBookPro laptop, Late 2013
OS: Mac OSX 10.9.4
GPU: NVIDIA GeForce GT 750M 2048 MB
Harddisk: APPLE 1TB SSD
CPU: 2.6 GHz Intel Core i7
Memory: 16 GB 1600 MHz DDR3
Resolume: 4.1.11

Setup: laptop as control monitor, 1 fullscreen output at 1920x1080 on HDMI output

Benchmarks 1920x1080
Clean: 13 layers
Noise: 10 layers

Benchmarks 640x480
Clean: 20+ layers at (47fps)
Noise: 20+ layers at (37fps)

Joris, any chance you guys could make a 4K benchmark file as well?

[clanoue]

Model: Apple MacBoor Pro. Late 2013
Operating System: OSX 10.9.4
Graphic card: GT 750 M 2GB
Harddisk: 512GB PCI-E SSD
Processor: 2.3GHZ i7
Memory: 16GB ddr3 1600mhz
Resolume: 4.1.11

Output Setup: 1920x1080 via mini display port to VGA adaptor

Benchmark for 1920x1080
Clean content: 8 at 30 FPS
Noise content: 7 at 30 FPS

Benchmark for 640x480
Clean content: 20 at 34-42FPS
Noise content: 20 at 34-42FPS

[0x80]

I've done a slightly different benchmark, because I am interested to know how the 750M card and the integrated Iris Pro graphics compare in terms of video performance. Mavericks won't let you switch the GPU manually anymore in its preferences, but I found this app called gfxCardStatus which seem to let you do it. You can't switch off the discrete graphics card if you have an external monitor attached, so that's why I left out the 2nd display measurements for the Iris graphics. Also I didn't test 640x480 since I figured it doesn't help for this comparison.


Model: MacBookPro laptop, Late 2013
Operating System: Mac OSX 10.9.4
Graphic card: NVIDIA GeForce GT 750M 2048 MB and Intel Iris Pro
Harddisk: 512Mb PCI-e Flash
Processor: 2.3 GHz Intel Core i7
Memory: 16 GB 1600 MHz DDR3
Resolume: 4.1.11

Only benchmarks for 1920x1080, but with different GPU and output configurations.

----
GeForce 750M. Fullscreen output on 2nd display at 1920x1080

Clean content: 9 layers
Noise content: 9 layers

----
GeForce 750M. Fullscreen output on laptop at 1440x900

Clean content: 9 layers
Noise content: 9 layers

----
Iris Pro. Fullscreen output on 2nd display at 1920x1080

Clean content: N/A (no external display when switched to integrated)
Noise content: N/A (no external display when switched to integrated)

----
Iris Pro. Fullscreen output on laptop at 1440x900

Clean content: 9 layers
Noise content: 9 layers

A few remarkable observations:
* Quite a bit slower than the benchmark of Oaktown, although the only difference is 2.3 vs 2.6gHz cpu.
* Outperformed by the MackBookPro Retina Mid 2012
* Different output and GPUs setups didn't really make a difference.

[Oaktown]

0x80, this is all great information but what's the point of testing a system without external displays? Testing your system without any external display would be like taking a Ferrari for a test drive while it's on 4 jacks. Unless you're going to run a show on your 15" monitor, you need to use external displays.

If you want to test the Intel Iris Pro Graphics with external displays on a Macbook Pro, you need to use a lower end Macbook Pro without a discrete GPU. But if you have a Macbook Pro with a discrete GPU (GT 750M in your case), you'll need to use that card to output graphics to your external displays and that's a good thing. Trust me, there will be a big difference between the two systems especially as you try scaling up.

[0x80]

To avoid polluting this benchmark thread with discussions I've responded in this thread:

viewtopic.php?f=12&t=11301&p=45353#p45353