Test 1, 8 seconds
Test 2, 1m 15s
Test 3, 4m 1s
Test 4, 4m 19s
Test 5, 4m 58s
New Mac Pro (Late 2013)
CPU: 3.5 GHz, 6-Core Intel Xeon E5
Memory: 32GB - From Apple
Storage: 1TB SSD - From Apple
Graphics: AMD FirePro D500 3072MB
Lightroom 5.3
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Lightroom 'Test Mule' on the nMP
- Thread starter Alchemist
- Start date
- Sort by reaction score
Test 1, 8 seconds
Test 2, 1m 15s
Test 3, 4m 1s
Test 4, 4m 19s
Test 5, 4m 58s
New Mac Pro (Late 2013)
CPU: 3.5 GHz, 6-Core Intel Xeon E5
Memory: 32GB - From Apple
Storage: 1TB SSD - From Apple
Graphics: AMD FirePro D500 3072MB
Lightroom 5.3
Man, I am so close to that 6core Mac Pro in this particular test.. I bet If I had my ram disk enabled my scores might be better lol but it's cheating...
Great scores. I wonder if the quad core Mac Pro with 1TB SSD could crack the 4m barrier
Great work!
Suggestion: You might consider moving (or replicating) the 5 test results columns to the first few columns of the spreadsheet. I have a 30" monitor and it was tough to view what mattered most (system and times) even on this big screen.
You might also consider highlighting the best score for each test.
Thanks for your efforts and those who contributed!
Suggestion: You might consider moving (or replicating) the 5 test results columns to the first few columns of the spreadsheet. I have a 30" monitor and it was tough to view what mattered most (system and times) even on this big screen.
You might also consider highlighting the best score for each test.
Thanks for your efforts and those who contributed!
Thanks to everyone who has taken the time to run the tests. I've still got results to add, so bear with me. I've modified the spreadsheet a touch to make it clearer to read and will keep working on that. I'm quite busy at the moment so getting it done when I can!
Test Mule - Benchmarking Spreadsheet
Keep sending me the results folks!
Test Mule - Benchmarking Spreadsheet
Keep sending me the results folks!
Tested the MP3,1 and late 2013 iMac in preparation for the nMP
Mac Pro 3,1 (early 2008) 2x2.8 Xeons, 16GB 800 MHz DDR2, ATI 5870 1GB, Samsung 840 Pro 256GB (backplane is SATAII) running OSX 10.8.5 & LR 5.3.
Test 1 - 0m28s
Test 2 - 2m24s
Test 3 - 8m04s
Test 4 - 7m55s
Test 5 - 7m28s
iMac 27" (late 2013) core i7 3.5, 16GB 1600 DDR3, NVIDIA GTX 780M 4GB, Apple/Samsung PCIe SSD 512GB running OSX 10.9.1 & LR 5.3.
Test 1 - 0m10s
Test 2 - 1m43s
Test 3 - 4m50s
Test 4 - 4m47s
Test 5 - 4m48s
Awaiting a 6c nMP for testing. Figured I'd record these of my current systems for comparison, but it also might help someone out who is comparing the late-2013 iMac vs the nMP.
Mac Pro 3,1 (early 2008) 2x2.8 Xeons, 16GB 800 MHz DDR2, ATI 5870 1GB, Samsung 840 Pro 256GB (backplane is SATAII) running OSX 10.8.5 & LR 5.3.
Test 1 - 0m28s
Test 2 - 2m24s
Test 3 - 8m04s
Test 4 - 7m55s
Test 5 - 7m28s
iMac 27" (late 2013) core i7 3.5, 16GB 1600 DDR3, NVIDIA GTX 780M 4GB, Apple/Samsung PCIe SSD 512GB running OSX 10.9.1 & LR 5.3.
Test 1 - 0m10s
Test 2 - 1m43s
Test 3 - 4m50s
Test 4 - 4m47s
Test 5 - 4m48s
Awaiting a 6c nMP for testing. Figured I'd record these of my current systems for comparison, but it also might help someone out who is comparing the late-2013 iMac vs the nMP.
Thank you so much for the iMac numbers. That machine is fast, still deciding between your configuration (except with a 1TB SSD) or the mac pro.
no problem. i got the iMac as an interim solution to hold me over until the nMP was available, with the possibility that it would fulfill or exceed all my needs (i'd then keep it and cancel the nMP order).
it's a really, really nice machine. it may not suffice for those doing major video processing (that's not me) or for someone who cannot and/or refuses to make do with the display, but it sure does have quite a bit going for it.
by my estimates, the nMP is about $1,800 more for a similar configuration when one adds in the cost of a thunderbolt display. that may be an appropriate context in which to view any performance gains.
The thing killing me is that I already have an apple 30" Cinema Display that I could use if I got a Mac Pro or sell if I got the imac.
you could use it alongside the iMac, too.
i have two 23" ACDs (matte) and an LED ACD. whether i stick with the iMac or go to the nMP i want to consolidate with a 2-monitor setup.
ps i ran the photoshop speed test and posted in that thread if it is of interest to you.
Something or someone killed the spreadsheet. I've restored from a backup and added all the new results I have received. Latest below. The 8 core Mac Pro appears to be a real beast.
Latest Test Mule Spreadsheet
Latest Test Mule Spreadsheet
I think the pcie ssd really help to boost the spped of the 8-core. The xeon 2667 v2 might give slightly better result
I also wonder how many cores/threads were actively used when the test was done in the 8-core.
Lastly what does FM mean? Thanks
Last edited:
As the spreadsheet has been corrupted once already, I thought it might be useful, to have a centralised link for all things relating to this test. I've placed all the relevant posts on my blog under a single unifying tag so new contributors can find everything in one place.
Lightroom Test Mule - Collated Info
Lightroom Test Mule - Collated Info
The Ars Technica review of the nMP is up and it's a very interesting read that offers another perspective. It's mainly aimed at other sectors of the pro user market rather than photographers, but I was excited and surprised to find that my Lightroom testing is mentioned on page 6! For those of you who haven't seen the article yet, here it is:
Ars Technica - Mac Pro Review
I'm still on the lookout for results for the Test Mule benchmark, specifically from anyone else who may have a nMP. It'd be particularly interesting to see results from a 4c model or a 12c and additional 6/8c models.
Ars Technica - Mac Pro Review
I'm still on the lookout for results for the Test Mule benchmark, specifically from anyone else who may have a nMP. It'd be particularly interesting to see results from a 4c model or a 12c and additional 6/8c models.
had a chance to run this on the nMP i received yesterday.
nMP 6c, 12GB, D500, 512 running OSX 10.9.0 & LR 5.2
Test 1- 0m7.5s (2.5s faster than the 2013 iMac = 25%)
Test 2 - 1m35s (8s faster = 8%)
Test 3 - 4m03s (47s faster = 16%)
Test 4 - 4m09s (48s faster = 17%)
Test 5 - 4m11s (37s faster = 13%)
i have 4x16GB chips sitting here. i forgot to remove the 3x4 and pop in 1x16 to make it closer to apples-to-apples. perhaps i will re-run with 1x16 or with 4x16 later on. don't know if any of these tests are memory-constrained.
geekbench3 (64-bit) of the imac vs nMP is 3940/14968 vs 3600/20300 single-core/multi-core.
realized i ran the tests on 10.9.0. not sure if it makes a difference vs 10.9.1 but i might retest.
Last edited:
Updated to 10.9.1. Geekbench went from 3600/20300 to 3617/20324.
Tests 1, 3, 4 and 5 had basically the same results but Test 2 (Import) dropped from 1m35s to 0m57s. I re-ran that test a few times (with new catalogues) and even saw some faster results.
Additional updates:
Re-ran tests with 32GB installed (2x16) and observed improvement in Tests 2. Geekbench=3616/20022
Test 1: 0m7.5s (no change)
Test 2: 0m57s (38s drop vs 12GB)
Test 3: 4m02s (no change)
Test 4: 4m10s (no change)
Test 5: 4m13 (no change)
Re-re-ran tests with 64GB installed (4x16). Geekbench=3654/21081
Test 1: 0m7.5s (no change)
Test 2: 0m50s (7s faster vs 32GB)
Test 3: 3m52s (10s faster vs 32GB)
Test 4: 4m03s (7s faster vs 32GB)
Test 5: 4m09s (4s faster vs 32GB)
To sum up, the 6c nMP/d500 with 64GB RAM vs 2013 iMac i7 w/16:
Test 1: 7.5s vs 10s (25% redux)
Test 2 - 0m50s vs 1m43s (51% redux)
Test 3 - 3m52s vs 4m50s (20%)
Test 4 - 4m03s vs 4m57s (18%)
Test 5 - 4m09s vs 4m48s (14%)
I ran all tests multiple times to confirm.
Last edited:
i updated the spreadsheet. the top 3 rows (current fastest times) are from my 6-core nMP. included the results from 3 memory configurations.