Used for storage, signal processing and analyses of psychophysiological data, primarily MRI data. I originally used a GIGABYTE X399 AORUS Xtreme motherboard but the TR4 socket pins bent during installation. I believe this resulted from a flawed CPU socket and not an installation mistake; the Foxconn TR4 socket was flimsy and very hard to secure after closing. I discovered the bent socket pins after hours of troubleshooting poor RAM performance (only 3 sticks were recognized at a time and two slots appeared dead). Nonetheless, Amazon worked with me and provided a partial refund, which was enough to purchase a ASRock X399 Taichi motherboard, which was a breeze to install and has given me zero problems so far. Although, admittedly, I've use this workstation for any demanding work yet.
I installed Ubuntu 18.04 on the SSD, which is my primary OS, and Windows 10 on the SATA for occasional MS Office use and for use with statistical packages that don't run very well on Linux (e.g., SPSS) or aren't available on Linux (e.g., AMOS). I've thus far installed NeuroDebian(http://neuro.debian.net/) and all packages run smoothly. I can't wait to try out parallelization with AFNI's. I was unable to run the most recent version of MRIcroGL due to my GPU's use of the newer OpenGL (4.4); the current MRIcroGL appears to only run on OpenGL 3.3 but I, for some reason, was able to successfully run an older version of MRIcroGL and MRIcron.
I plan to update my description after I've had a chance to works with some MRI data and have a better idea about how well I am able to take advantage of the multicore CPU. Sorry for the bad pic. It's all I have right now. I'll post something a little nicer after I up my zip tie game.