A small scientific workstation for the home office. The parts are not in the top tier but relatively new ones. So far I am very pleased.
Intel Core i5-9600K is the new mid-tier processor with the soldering, which is good for the heat spreading. Although AMD Ryzen 2600X or 2700X is more cost-effective, I choose Intel processor because some scientific software is optimized with Intel Math Kernel Library and is much faster.
I am satisfied with the quiet but powerful Corsair H100i PRO.
Gigabyte Z390 I AORUS PRO WIFI is the lastest Mini ITX board for Intel. It gives some potential to overclock the i5-9600K. However, I haven't really pushed the limit yet.
With XMP technology, it is very easy to boost the Corsair Vengeance LPX 2 x 16GB DDR4 to 3200MHz.
In consideration of the SSD endurance, I choose Samsung 970 Pro over 970 EVO. An old Samsung 850 Pro 256 GB is used for Ubuntu.
NVIDIA GeForce RTX 2070 Founders Edition is a little bit over-priced now. I still decide to buy it because I want to try the Tensor Cores and DLSS for the deep learning. I have to say that NVIDIA Deep Learning SDK is much more advanced than AMD one.
The Metallic Gear NEO MINI case is somehow large but good for the PC temperature. You can tell it from the combined CPU & GPU stress test result. The installation process is quite straightforward.
SeaSonic FOCUS Plus Platinum 550W gives enough watts for a stable running.