The workstation build for Machine Learning in 2023
The PC build that I decided to use for my research related to machine learning and generative adversarial networks (GAN) can be found here. https://pcpartpicker.com/list/vb6Ft7
This build is more of a budget-friendly build for a machine learning workstation with the newest CPU and GPU in the market. This blog does not contain technical performance testing, but more from what I learn from various online videos and articles. It provides the stuff that you should be careful of when building ur own machine-learning workstation.
The specs included:
CPU - AMD Ryzen 9 7950X3D 4.2 GHz 16-Core Processor
CPU Cooler - NZXT Kraken Elite 360 RGB 78.02 CFM Liquid CPU Cooler
Motherboard - Asus ROG STRIX X670E-E GAMING WIFI ATX AM5 Motherboard
RAM - TEAMGROUP T-Force Delta RGB 128 GB (4 x 32 GB) DDR5-6000 CL38 Memory
SSD - Samsung 990 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive * 2
GPU - MSI SUPRIM LIQUID X GeForce RTX 4090 24 GB Video Card * 2
Computer Case - Lian Li PC-O11 Dynamic ATX Full Tower Case
Power Supply - EVGA SuperNOVA 1600 P2 1600 W 80+ Platinum Certified Fully Modular ATX Power Supply
Fans - Lian Li UNI FAN SL V2 64.5 CFM 120 mm Fans 3-Pack
CPU
As for the CPU, I chose the AMD Ryzen 9 7950X3D. What so special about this CPU is that is it provides great performance at a very optimized temperature and powerful consumption, which is very suitable for this build, considering we have a dual RTX 4090 graphics card in our build, which will generate more heat and consume of power. This CPU can create that balance for the workstation, as it can run for a long time; power consumption and temperature might be issues that need to be considered.
Motherboard
The motherboard that will be used is the ASUS ROG STRIX X670E-E Gaming. Since RTX 4090 is still a very powerful graph card and pretty new as well. A lot of consumers might not know how to put 2 4090 into their PC, There are not that many online videos that talk about how to fit 2 4090 into their motherboard. As a normal consumer, I studied a lot from videos and read many articles to identify what I needed to power this dual 4090 setup. I came across a few motherboards to find out the X670E-E is the one that is able to fit two 4090, but without sacrificing the performance. The X670E-E contains 2 x PCIe 5.0 x16, which, unlike most of the motherboards out there, only provides 1 x PCIe 5.0 x16, and 1 x PCIe 5.0 x8, which will significantly lower the performance of the second graphics card. The motherboard also provides up to 128GB of RAM and supports DDR5.
GPU
The dual RTX 4090 GPU that will be used in this workstation is the MSI SUPRIM LIQUID X GeForce RTX 4090. The main reason that I chose this GPU is because of its size. As you might know, the 4090, as the newest and powerful graphics card for normal consumers, is much bigger than its previous generations. There are a lot of cases in which people find it hard to find a motherboard or even a computer case to fit those Graphics cards. However, what is so unique about the MSI SUPRIM LIQUID X GeForce RTX 4090 is a liquid-cooled GPU. It is smaller than the traditional GPU that is cooled down directly by the fans, which is a lot bigger for the size. Also, one thing that is worth pointing out. When you have two 4090 installed in the motherboard, you should also consider the gap between them. In a lot of cases, you might find yourself barely able to fit 2 4090 into the motherboard, but the problem is that you might block the air flows between the two graph cards, which might significantly lower the performance or overheate the graph cards. Therefore, I choose MSI SUPRIM LIQUID X GeForce RTX 4090, as it is smaller, water-cool, and provide more space in between the graphics card to enhance airflow.
Subscribe to my newsletter
Read articles from Archie Tan directly inside your inbox. Subscribe to the newsletter, and don't miss out.
Written by
Archie Tan
Archie Tan
I'm Archie, a Computer Science undergraduate student and a researcher. With my passion for programming and desire to raise the bar for everyday life, I like learning new skills in my free time and reading books. I'm currently addicted to learning AI/ML and Computer Vision.