↝ OWC / MacSales.com...
↝ diglloyd B&H Deal Finder...
Buy other stuff at Amazon.com...
2017 iMac 5K: diglloyd Photoshop Benchmarks
Related: 2015 iMac 5K, 2017 iMac 5K, 4K and 5K displays, computer display, GPU, iMac, iMac 5K, Mac Pro, Macs, memory, Photoshop, SSD, video
MPG recommends 64GB OWC memory (big savings!) in: 27" Retina 5K Display, 4.2 GHz Intel Core i7 (Skylake), 1TB Flash, 8GB memory, AMD Radeon Pro 580 Graphics Card (8GB).
The machines tested here are the fastest possible CPUs and GPUs as of mid 2017.
A note on the 8-core 3.3 GHz Mac Pro
These tests use a 2013 Mac Pro 3.3 GHz 8-core, the fastest Mac Pro available, and one that Apple does not sell—it needs to be an upgraded CPU. Apple offers a 3.0 GHz 8 core (slower than 3.3 GHz), and it would be a bit slower than the 3.3 GHz 8-core Mac Pro tested here.
Summary: the 2017 iMac 5K offers a substantial speed boost over its 2015 predecessor and boots the fastest possible Mac Pro into 2nd place. However, it cannot beat the fastest Mac Pro.
The diglloyd Speed1 photoshop benchmark uses a mix of the most commonly used Photoshop operations with a file size that allows everything to stay in memory. Hence it accurately represents what one might expect in everyday use of Adobe Photoshop CC. Specialty operations such as Liquify and other GPU-intensive tasks are just that—specialty, and are not included in this suite.
While the 4.2 GHz CPU of the 2017 iMac 5K is only 5% faster than the 4.0 GHz CPU, 2400 MHz memory speed is also quite a lot faster than 1866 MHz memory (28% faster), and the GPU is considerably more capable—apparently these things add up for Photoshop use.
The diglloyd Medium photoshop benchmark uses a mix of the core Photoshop operations with a moderately larger file size that exceeds what most users are likely to use, using about 15GB of memory in Photoshop.
The 3.3 GHz 8-core Mac Pro trounces even the 2017 iMac 5K. Apparently the latest Photoshop and macOS have allowd the D700 GPUs to run closer to their potential.
The 2017 iMac 5K handily outperforms the 2015 iMac 5K by about 14%, which is notably more responsive (MPG deems 10% the threshold for a noticeable difference).
The diglloyd Large photoshop benchmark uses a mix of the core Photoshop operations with a large file size that exceeds what most users are likely to use, using about 30GB of memory in Photoshop.
The 2017 iMac 5K matches the 2013 Mac Pro. This seemed odd, but repeating the test (with 20 iterations each time) showed the same results.
The 2017 iMac 5K has an SSD that is at least 2X faster than the Mac Pro, so that it can operate the CPU and GPU with less waiting on disk I/O. Even though this test fits into 64GB memory quite easily, Photoshop has some behavior that may still hit the disk.
The 2017 iMac 5K is 15% faster than the 2015 iMac 5Kon this test.
The diglloyd Huge photoshop benchmark uses a mix of the core Photoshop operations with a file size that requires about 56GB of memory usage in Photoshop, well beyond what most users are likely to encounter. Due to memory limits and the resulting variability in runtime, it is advised to run 10 iterations or more of this test.
The Mac Pro assets itself as the leader, but note that the 2017 iMac 5K is not far behind, only about 7% slower. That’s because some disk I/O is incurred, and the 2017 iMac 5K has an SSD that is at least 2X faster, so that it can operate the CPU and GPU with less waiting on disk I/O.
The 2017 iMac 5K has a faster CPU and GPU and SSD than the 2015 iMac 5K, and this all comes together to decisively outperfom the 2015 model by 22%.