Thank you for buying your 2017 iMac 5K and/or iMac Pro at B&H Photo, and for upgrading it at OWC MacSales.com. Both vendors made and make my extensive tests of the Apple iMac Pro and other Macs and gear possible.
Not sure which Mac you need, or how to configure it, plan for backup, etc? Consult with Lloyd.
I am back home from my Jan 2 - Jan 5 trip to OWC labs in Woodstock, IL. It was a tough one, with not a lot of sleep and about 40 hours of testing. Plus some more testing upon my return, to complete the comparisons with other machines.
Thank you for buying from OWC / MacSales.com which hosted me at OWC labs and made machines available for testing, and buying iMac Pro from B&H Photo, which sent me an iMac Pro which is sitting on my desk for a few weeks. This research was a huge amount of work and is provided for all comers to read, so subscribing is much appreciated.
Photographers running macOS will want to follow my blog posts and review of the iMac Pro at MacPerformanceGuide.com very carefully for the next few days.
A lot of stuff will be appearing, including the most in-depth and useful review of the 2017 iMac Pro available anywhere on the internet, at least for still photographers and anything similar. By that I mean whether it will help or not help your own photographic and general computation-intensive work, and whether the iMac Pro is a needless cost or a win. It all depends.
Wide-ranging tests will start appearing shortly, where I might save readers thousands of dollars by showing the value proposition of the iMac Pro versus the 2017 iMac 5K. However, the good news is that Adobe Lightroom users will find joy, if the price of the iMac Pro can be swallowed.
There are also compatibility issues with the new iMac Pro (and 2013 Mac Pro) with both macOS 10.13.2 and 10.12.6 (latest versions with the security updates).
It is my suspicion that Apple’s claims of fixes for Meltdown and Spectre are only half the picture and might be misleading, since I am observing performance issues in my own software whose code has always been highly performant, for 10 years. Now it’s up to 3X slower for no reason (no changes to the source code). And unless carefully tuned, there is a 2X to 3X variation in performance with only small changes in buffer size and thread count. Worse, that tuning changes by machine and drive speed. Over a decade, I never saw such extreme behavior—it all seems to have started with the recent security updates to 10.13.2 and 10.12.6.