I’ve been planning this system for a while. The idea is simple: build one powerful PC that can handle both local AI work and high-end gaming without feeling outdated too quickly.
This won’t be just a gaming PC. I also want to use it for things like Ollama, Open WebUI, ComfyUI, image generation, local LLM testing and future AI experiments.
At the same time, I still want strong 1440p / 4K gaming performance with high settings and ray tracing where possible.
Main goal
The main goal is to build a stable, powerful and long-term system for:
- Local AI tools
- Gaming
- Image generation
- AI experiments
- Development work
- Heavy multitasking
- Future upgrades
I didn’t want to rush the whole build at once, so I decided to collect the parts step by step.
Why I picked NVIDIA
For local AI work, the GPU matters a lot.
After looking at the current AI software ecosystem, NVIDIA still feels like the safest and most practical choice for my use case. CUDA support and better compatibility with many AI tools make things much easier.
This is especially important for tools like:
- Ollama
- Open WebUI
- ComfyUI
- Stable Diffusion workflows
- Local LLM tests
- AI image generation
- Future automation projects
Current parts list
GPU: MSI GeForce RTX 5070 Ti 16G Gaming Trio OC
CPU: AMD Ryzen 9 9950X
Motherboard: ASUS ROG Strix X870E-H Gaming WiFi
RAM: Kingston Fury Beast RGB DDR5 64GB (2×32GB) 6000MHz CL36 EXPO
SSD: Samsung 9100 PRO 2TB Gen5 NVMe
Case: Thermaltake The Tower 600 “Bumblebee”
CPU Cooler: ASUS ROG Ryujin III 360 ARGB Extreme
PSU: Corsair RMe Series RM1200e 1200W 80+ Gold ATX 3.1 / PCIe 5.1
CPU choice: why Ryzen 9 9950X?
At first, I was planning to use a Ryzen 9 7950X3D. It is an excellent gaming CPU, but my build is not only focused on gaming.
I also care about AI workflows, productivity, multitasking and long-term performance. After a bad purchase experience with a supposedly “new sealed” 7950X3D, I returned it and changed direction.
The Ryzen 9 9950X made more sense for this build because I wanted stronger all-around performance, not only gaming-focused performance.
Gaming target
For gaming, I’m aiming for:
- 1440p / 4K gaming
- High or ultra settings
- Ray tracing where possible
- Smooth performance in modern games
- A system that can stay strong for years
AI target
For AI, I want to test and run:
- Local LLMs
- Ollama models
- Open WebUI
- ComfyUI workflows
- AI image generation
- AI translation experiments
- Personal AI assistant ideas
- Local automation projects
Build direction
This is not meant to be a budget build. I want something powerful, stable and flexible enough for both gaming and AI work.
I’ll keep updating this topic as the build progresses, including photos, changes, tests, temperatures and real-world performance results.
What do you think?
Would you build something similar for a mixed AI + gaming PC?
Would you change the GPU, CPU or platform for local AI workloads in 2025?
