↝ OWC / MacSales.com... ↜
↝ diglloyd Deal Finder... ↜
Buy other stuff at Amazon.com...
2016 MacBook Pro: Adobe Lightroom Import RAW Files with 1:1 Previews
Related: 2016 MacBook Pro, GPU, laptop, Lightroom, Mac Pro, MacBook, MacBook Pro, Macs, Photoshop, RAW, SSD, video
MPG tested a fully-loaded 2016 MacBook Pro with 2TB SSD.
This test is as real world as it gets and for a very large number of users: importing digital camera RAW files into Adobe Photoshop Lightroom.
How does a 3-year-old 2013 MacBook Pro beat out the fastest possible build of the 2016 MacBook Pro*? These tests were (tediously) repeated three times for each machine**, with the best result from each shown (consistency was strong). The explanation is that 2016 MacBook Pro performance declines with sustained usage.
It is pathetic that the 2016 MacBook Pro runs at 2.9 GHz and with a 4.5X faster SSD, and the 2013 model runs at 2.6 GHz and yet the 2013 MacBook Pro wins.
Similarly, how does an 8 core Mac Pro barely beat a 4 core iMac 5K? It is a testament to poor processing algorithms that Lightroom does so poorly on the 8 core Mac Pro, which has twice the CPU cores. The reason is that Lightroom imports files serially (one after another), instead of handling one file per CPU core with asynchronous disk I/O. While LR uses multiple CPU cores on each file, it cannot use the CPUs fully most of the time. It is a classic case of serialized computationfor no good reason when highly parallel computation is possible—poor software engineering. There is no substitute for parallelism across the batch of files as a whole.
* The MPG 2013 Mac Pro with dual D700 GPUs, 64GB memory, 1TB SSD, and custom upgrade to the fastest Intel Xeon chip (3.3 GHz 8-core) still is beaten by the late 2015 iMac 5K, but it does beat out the two MacBook Pros. It has 8 CPU cores, but Photoshop uses at most 2-3 cores for this job, so the slower CPU speed means the Mac Pro loses out.
** New catalog each time, identical configuration of Lightroom and catalog prefs, system observed to be quiescent before starting each test.
Test machine notes
All systems are/were the fastest available in their category. All running macOS Sierra 10.12.1.