Since few months I was thinking of building a dedicated, no-compromise machine for Artificial Intelligence and Machine Learning. The goal was to build a workstation which can be used to have consistent throughput for hours-long training sessions, massive dataset processing, and future-proofing for the next generation of large language models (LLMs) keep a check on the budget. Every component choice was well thought through, prioritizing stability, thermal management, and raw compute power.
With the budget constrains, I knew I had to use AMD Ryzen 9 9950X and NVIDIA GeForce RTX 5090 with the consideration to upgrade to RTX Pro 6000 in future
Here is the component list:-
| Component | Selection |
|---|---|
| CPU | AMD Ryzen 9 9950X 16 Cores / 32 Threads (Zen 5) |
| CPU Cooler | ARCTIC Liquid Freezer III Pro 420 |
| Thermal Compound | Thermal Grizzly Kryonaut Extreme |
| Motherboard | Gigabyte X870E AORUS ELITE WIFI7 |
| Memory | Corsair Vengeance 96GB (2x48GB) DDR5-6000 CL30 |
| Storage | 1x4TB Samsung 9100 PRO |
| Video Card | Zotac GAMING SOLID GeForce RTX 5090 32 GB |
| Case | Fractal Design Meshify 3 XL |
| Power Supply | Corsair HX1500i 1500W, 80 PLUS Platinum, ATX 3.0 |
| Case Fan | ARCTIC P14 PST 72.8 CFM 140 mm Fans 5-Pack |
And some benchmarks
