While studying software development at SAIT, I learned how to build and repair computers in a computer hardware class. Ever since then I have been fascinated by computer hardware, and have created some dream PCs online using tools like PCPartPicker's System Builder tool. A few years after graduating, I was able to afford to build my own dream machine. Here are the details of my build, and why I made these decisions.
Motherboard
Asus ROG STRIX Z390-E GAMING
For this build it was important that the motherboard was able to push the limits of home computing. This choice was driven my ambitions to teach myself machine learning at home. Larger machine learning projects require a lot of RAM, and this board supports up to 4x DDR4 RAM Sticks with a maximum capacity of 64GB. This motherboard provides features that efficiently transfer heat away from the CPU and also has two M.2 slots allowing hard drives to be installed directly on the motherboard itself.
Processor
Intel Core i9-9900K
When it came to the processor I did not want to sacrifice performance. The Intel Core i9-9900K is an incredibly powerful processor that ships with 8 cores. This processor supports Intel Hyper-Threading technology, meaning 2 threads run on each core for a total of 16 threads. This processor uses its resources efficiently by increasing processor throughput, and improving overall performance on multithreaded software. The processor base frequency is 3.6 GHz but can boost 2 of it's cores up to an impressive 5 GHz.
GPU
Asus RTX 2070 Super
Performance for software development and content creation was the primary design factor for this build, and I wanted to include a high performance GPU in this list of components. The GeForce RTX line of GPU's by Nvidia are designed for high performance computing. They are not only strong for gaming, but are also excellent for professional media and 3D content creation. This GPU contains both Nvidia Tensor and RT cores, which accelerate deep learning models and real-time raytracing. This means faster media render times, and the ability to create deep neural networks in a fraction of the time it would take a CPU to develop.
Memory
32GB Corsair Vengeance Pro
Memory is not a bottleneck that I wanted to affect my work, so I ensured that it wouldn't be an issue by investing in 32GB. As mentioned in the GPU section, I have an interest in learning more about machine learning outside of web development. Machine learning requires large amounts of data and if that data is stored in memory, you better have the hardware to back it up. Machine learning frameworks may also duplicate all of the data stored in memory to perform specific operations. If you are working with 16GB of data, there may be a short period of time that 32GB of data in memory is required for the framework to perform it's computation.
Power supply
Corsair HX750i
Power supplies vary in performance, but I chose to go with a platinum grade supply to ensure that the system is provided with a steady stream of power. This 750W supply has no difficulty powering this system.
Case
NZXT H500i
This is the most beautiful case I could find without looking too flashy. It is black steel all around with a tempered glass panel on the side. The final build looks absolutely gorgeous on both the inside and outside of this case. It is also a part of the "i" series, which contains NZXT's "Smart Hub". This allows you to program things like the case LED colours and fan speeds. The LEDs even support animations, and colour can be controlled by measuring the temperature of the CPU. I am in love with this case.
Storage
2X Western Digital 250GB M.2 SSD
One design goal of this machine was to have a dual boot configuration. One drive for Windows 10, and one drive for Ubuntu Linux. Yes, I could just run Linux in a virtual machine, or even through the somewhat new Windows Subsystem for Linux, but I wanted to run a full Linux distro on bare metal, without any virtualization.
1X Seagate 2TB HDD
This is way more than enough to get the build started and working. All of the applications are stored on the M.2 SDDs, but the rest of my documents and media are stored on this 2TB HDD.
Liquid cooling
NZXT Kraken X62
The Intel Core i9 can run really hot when running at full capacity. I chose to use liquid instead of air in order to achieve maximum cooling efficiency. It also just looks really nice inside of this build, and includes a LED ring with programmable colours and animations.