Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

'Real World' Photoshop Benchmark - Speed Test

Killerbob said:
I like the fact that everyone is saying how little the nMP does to photography etc., and how it is only a video editing powerhouse, and then I see the first score from an MP6,1... 1.8sec



That basically means it is by far the fastest m.....f..... I have seen! It is 10 times faster than my MP3,1/2.8Octo/16GB/Accelsior/GTX680:eek:
Click to expand...


I notice that nMP has 64gb of ram! That's bound to help! But yeah I concur with your notion it's fast!
 
I am sure the 64GB RAM plays a role, but I was always afraid that us photogs wouldn't see a marked improvement moving to the nMP. I already have a decent videocard (GTX680) and great storage (PCIe Accelsior), so I felt it wouldn't much of an upgrade. Now I see a marked PS improvement, I'll start planning for when, not if...

I also notised that the nMP was "only" equipped with D500s, and I would certainly get the D700s (for gaming in bootcamped Windows). Probably I would stick to 32GB RAM, but I think that's enough...
 
jerryrock said:
MacPro 6,1 3.5GHz 6-Core Intel E-5, dual AMD FirePro D500 (Photoshop only uses one), 64GB DDR3 1866MHz ECC Ram, 256GB Apple PCI SSD.

Photoshop CC (version 14.1.2 x64) OSX 10.9.1 (Mavericks)

Time 1.8 seconds
Click to expand...


Holy hell that is fast.

I wonder why VR got a time of 9 seconds using similar hardware (although with only 32 GB ram and CS6).
 
Mac Pro 5.1 Hex 3.33
HD 7970 reflashed
12GB 1333 Memory
Samsung Pro 512GB on Sata3
CS6

10.3 seconds

Rechecked this time with my camcorder to get an accurate timing

10.03 seconds
 
Last edited:
rMBP late 2013, 2.6 i7, 512 SSD, 16 gb, Mavericks

11.4 seconds (using Photoshop CC).

MP 5.1, 3.33 hex, mechanical disk, 24gb, Mavericks

10.3 (also with CC)

edited: I forgot to let PS have all the MP's memory.
 
Last edited:
nMP questions

1. How fast will a nMPro be running PS CS6/CC and LR5 in its base configuration ( Quad core Xeon 3.7Ghz with 12 GB RAM and the standard dual video cards in comparison to my 2007 V1.1 dual core dual chip 266Ghz Xeon MacPro running 9GB of DDR2 FB-DIMM RAM and a bUS speed od 1.33GHz?:confused:

2. Will the nMPro be able to drive my old 20" Cinema display? ( It of course is not set up for Thunderbolt.)

3. I have heard that the SSD and RAM are user upgradeable. Are the video cards upgradeable? I thought I had heard that there may be outboard video cards that could be used... but that was a while ago.

4.Can the nMacPro boot older versions of OSX from bootable external drives?

I have use mostly PS CS6/CC and LR5 with NIK plug-ins.
 
hassiman said:
1. How fast will a nMPro be running PS CS6/CC and LR5 in its base configuration ( Quad core Xeon 3.7Ghz with 12 GB RAM and the standard dual video cards in comparison to my 2007 V1.1 dual core dual chip 266Ghz Xeon MacPro running 9GB of DDR2 FB-DIMM RAM and a bUS speed od 1.33GHz?:confused:

2. Will the nMPro be able to drive my old 20" Cinema display? ( It of course is not set up for Thunderbolt.)

3. I have heard that the SSD and RAM are user upgradeable. Are the video cards upgradeable? I thought I had heard that there may be outboard video cards that could be used... but that was a while ago.

4.Can the nMacPro boot older versions of OSX from bootable external drives?

I have use mostly PS CS6/CC and LR5 with NIK plug-ins.
Click to expand...

Maybe you should start your own thread to get answers or look for the already existing ones.

Anyway, to quickly answer your questions:

1. More than twice as fast.

2. Yes.

3. When third parties build compatible GPUs, yes. As of now, there are none yet.

4. Not likely, maybe OS X 10.8 Mountain Lion might work, but only OS X 10.9 Mavericks has the proper drivers for all the hardware included in the 2013 Mac Pro.


My Hackintosh with
  • i7 3.5 GHz CPU
  • 32 GB RAM
  • GTX670 GPU
  • 840 SSD
running OS X 10.8 Mountain Lion and PS CC got it done in 11 seconds, without any of those adjustments or restarts the ReadMe asked for. Maybe I could get an additional two seconds out of that, but the 11 seconds is good enough for my needs.
 
2006 Mac Pro, 2 x 2.66, 12 GB RAM, SSD, Photoshop CC
28.9 seconds

Late 2013 Retina Macbook Pro, 2.3 Ghz, 16 Gig RAM, Photoshop CC
14.4 seconds
 
SimeoneSergio said:
Seriously jaw dropping.
Click to expand...

Has that result been duplicated yet? No offense to the person who posted it, but could there possibly have been a mistake? It is just so much faster than the quad core nMP results that it is hard to believe it is accurate. It doesn't seem like six instead of four cores could make that much difference. Unless it is the 64 gig of memory? Is this speed test really that memory dependent?
 
PmattF said:
Has that result been duplicated yet? No offense to the person who posted it, but could there possibly have been a mistake? It is just so much faster than the quad core nMP results that it is hard to believe it is accurate. It doesn't seem like six instead of four cores could make that much difference. Unless it is the 64 gig of memory? Is this speed test really that memory dependent?
Click to expand...

The other variable is the version of PS. The other tests on nMPs sofar have been with 5-6, and not the latest CC.
 
Speed

I would have to guess even if I go the base MacPro it will be a hell of a lot faster running PS CS6/CC and LR 5 than my 4 core 2.66 GH V1.2 MacPro from 2006/2007 running 9GB of RAM:D
 
PmattF said:
Has that result been duplicated yet? No offense to the person who posted it, but could there possibly have been a mistake? It is just so much faster than the quad core nMP results that it is hard to believe it is accurate. It doesn't seem like six instead of four cores could make that much difference. Unless it is the 64 gig of memory? Is this speed test really that memory dependent?
Click to expand...

I suspect the 1.8 second time is simply the result of the last action using the timing feature in Photoshop

My guess is that Jerryrock failed to time the test properly using a stopwatch or similar device.
 
Infocusinc said:
I suspect the 1.8 second time is simply the result of the last action using the timing feature in Photoshop

My guess is that Jerryrock failed to time the test properly using a stopwatch or similar device.
Click to expand...

I agree. Without confirmation from others who have a nMP with similar/same config the result is simply too good to be true. FYI, a dual E5 8 cores with 256GB ram and ample amount of SSD space (running Windows 8.1 :D) the speed was around 8 seconds.
 
Infocusinc said:
I suspect the 1.8 second time is simply the result of the last action using the timing feature in Photoshop

My guess is that Jerryrock failed to time the test properly using a stopwatch or similar device.
Click to expand...

You're exactly right, I used the Photoshop timing feature, which apparently does not work with this action. I questioned the accuracy of a test that relied on the error that could be introduced by human reaction reaction time and attempted to get around it.

A more accurate Photoshop test with a database of machines can be found here:

http://ksimonian.com/Blog/2010/02/24/improved-photoshop-benchmark-cpu-speed-test-for-both-mac-pc-free-radial-blur-filter-test/

My results for this Photoshop blur test are as follows: 14.5 seconds

Late 2013 MacPro 3.5 GHz Intel® Xeon Processor E5-1650 v2, OSX 10.9.1 (Mavericks), Photoshop CC ( CS6.1.2 )
Adobe Photoshop Version: 14.1.2 (14.1.2 20130923.r.427 2013/09/23:23:00:00) x64
Operating System: Mac OS 10.9.1
System architecture: Intel CPU Family:6, Model:62, Stepping:4 with MMX, SSE Integer, SSE FP, SSE2, SSE3, SSE4.1, SSE4.2, HyperThreading
Physical processor count: 6
Logical processor count: 12
Processor speed: 3500 MHz
Built-in memory: 65536 MB
Free memory: 43700 MB
Memory available to Photoshop: 62834 MB
Memory used by Photoshop: 70 %

I also did the Blackmagic Disk Speed Tests that are already recorded in this forum. If you have any other specific tests you would like performed please let me know.
 
Last edited:
jerryrock said:
A more accurate Photoshop test with a database of machines can be found here:

http://ksimonian.com/Blog/2010/02/24/improved-photoshop-benchmark-cpu-speed-test-for-both-mac-pc-free-radial-blur-filter-test/

My results for this Photoshop blur test are as follows: 14.5 seconds
Click to expand...

My MacPro5,1 6c(W3680)/48GB/512GB PCIe SSD/Quadro4000 with Photoshop CC 14.1.2 gives a time of 17.7 seconds. Compare to a Dell T7600 with dual 8c(E5-2680)/256GB/1TB PCIe SSD/Quadro K5000 the time was a shocking 6 seconds.

I notice radial blur, which is used in this test, uses all CPU cores. Another benchmark by diglloyd http://macperformanceguide.com/OptimizingPhotoshopCS6-Benchmarks.html, which mainly use smart sharpen, only use a portion of the cores. Thus PS performance really depend on your workflow (ie. if you use a lot of radial blur more/faster cores is better. on the other hand if you only use smart sharpen there is no obvious difference even between a 2010 Mac Pro and a dual CPU Dell).
 
Last edited:
@jerryrock; can you perform the PS test we use in this discussion as well? With about 750 records it is worth it for comparison reasons...
 
For the test used in this discussion:

MacPro5,1 6c(W3680)/48GB/512GB PCIe SSD/Quadro4000 with Photoshop CC 14.1.2 x64 - 10.53 seconds

For comparison only
Dell T7600 - dual 8c(E5-2680)/256GB/1TB PCIe SSD/Quadro K5000 running Windows 8.1 with Photoshop CC 14.1.2 x64 - 8.56 seconds


I really look forward to how the 8c nMP perform in PS
 
Killerbob said:
@jerryrock; can you perform the PS test we use in this discussion as well? With about 750 records it is worth it for comparison reasons...
Click to expand...

OK I ran it again using my iPhone as a timer this time!

1st run: 9.10 seconds

2nd run: 9:08 seconds

I believe the variable is due to reaction time hitting the stop button on the timer. Can't this test be rewritten to use Photoshop's built in timing function?

MacPro 6,1 3.5GHz 6-Core Intel E-5, 64GB 1866MHz DDR3 (OWC) RAM, dual AMD FirePro D500's, Apple 256GB PCI SSD.
Adobe Photoshop Version: 14.1.2 (14.1.2 20130923.r.427 2013/09/23:23:00:00) x64
Operating System: Mac OS 10.9.1
System architecture: Intel CPU Family:6, Model:62, Stepping:4 with MMX, SSE Integer, SSE FP, SSE2, SSE3, SSE4.1, SSE4.2, HyperThreading
Physical processor count: 6
Logical processor count: 12
Processor speed: 3500 MHz
Built-in memory: 65536 MB
Free memory: 62538 MB
Memory available to Photoshop: 62806 MB
Memory used by Photoshop: 100 %
Image tile size: 1024K
Image cache levels: 4
Display: 1
Main Display
Display Bounds: top=0, left=0, bottom=1050, right=1680
Display: 2
Display Bounds: top=-1178, left=-120, bottom=0, right=1800
OpenGL Drawing: Enabled.
OpenGL Allow Old GPUs: Not Detected.
OpenGL Drawing Mode: Advanced
OpenGL Allow Normal Mode: True.
OpenGL Allow Advanced Mode: True.
NumGPUs=1
gpu[0].OGLVersion="2.1"
gpu[0].MemoryMB=3072
gpu[0].RectTextureSize=16384
gpu[0].Renderer="AMD Radeon HD - FirePro D500 OpenGL Engine"
gpu[0].RendererID=16915457
gpu[0].Vendor="ATI Technologies Inc."
gpu[0].VendorID=4098
gpu[0].HasNPOTSupport=1
gpu[0].CompileProgramGLSL=1
gpu[0].TestFrameBuffer=1
gpu[0].OCLSupported=1
gpu[0].OCLVersion="1.2 (Oct 1 2013 19:04:04)"
gpu[0].CUDASupported=0
gpu[0].glGetString[GL_SHADING_LANGUAGE_VERSION]="1.20"
gpu[0].glGetProgramivARB[GL_FRAGMENT_PROGRAM_ARB][GL_MAX_PROGRAM_INSTRUCTIONS_ARB]=[16384]
gpu[0].glGetIntegerv[GL_MAX_TEXTURE_UNITS]=[8]
gpu[0].glGetIntegerv[GL_MAX_COMBINED_TEXTURE_IMAGE_UNITS]=[16]
gpu[0].glGetIntegerv[GL_MAX_VERTEX_TEXTURE_IMAGE_UNITS]=[16]
gpu[0].glGetIntegerv[GL_MAX_TEXTURE_IMAGE_UNITS]=[16]
gpu[0].glGetIntegerv[GL_MAX_DRAW_BUFFERS]=[8]
gpu[0].glGetIntegerv[GL_MAX_VERTEX_UNIFORM_COMPONENTS]=[4096]
gpu[0].glGetIntegerv[GL_MAX_FRAGMENT_UNIFORM_COMPONENTS]=[4096]
gpu[0].glGetIntegerv[GL_MAX_VARYING_FLOATS]=[128]
gpu[0].glGetIntegerv[GL_MAX_VERTEX_ATTRIBS]=[16]
gpu[0].extension[AIF::OGL::GL_ARB_VERTEX_PROGRAM]=1
gpu[0].extension[AIF::OGL::GL_ARB_FRAGMENT_PROGRAM]=1
gpu[0].extension[AIF::OGL::GL_ARB_VERTEX_SHADER]=1
gpu[0].extension[AIF::OGL::GL_ARB_FRAGMENT_SHADER]=1
gpu[0].extension[AIF::OGL::GL_EXT_FRAMEBUFFER_OBJECT]=1
gpu[0].extension[AIF::OGL::GL_ARB_TEXTURE_RECTANGLE]=1
gpu[0].extension[AIF::OGL::GL_ARB_TEXTURE_FLOAT]=1
gpu[0].extension[AIF::OGL::GL_ARB_OCCLUSION_QUERY]=1
gpu[0].extension[AIF::OGL::GL_ARB_VERTEX_BUFFER_OBJECT]=1
gpu[0].extension[AIF::OGL::GL_ARB_SHADER_TEXTURE_LOD]=1
Serial number: XXXXXXXXXXXXXXXXXXX
Application folder: /Applications/Adobe Photoshop CC/
Photoshop scratch has async I/O enabled
Scratch volume(s):
Promise Pegasus, 5.46T, 4.34T free
Macintosh HD, 233.0G, 184.0G free
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back