I'm gearing up to dive into AI projects and I'm currently building my system. Here's what I've got so far:
- CPU: Ryzen 7 5700G
- Motherboard: B550
- Power Supply: 500W (planning to upgrade soon)
- RAM: 16GB 3600MHz
- Storage: 256GB NVMe, 128GB SSD, and 512GB HDD
My main question is about the graphics card—I'm on a budget that will likely get me either an RTX 3060 or an RTX 3060 Ti. I hear 8GB of VRAM is common, but is that enough for a beginner? I'm getting mixed advice, especially since local pricing is crazy (like $350 for a used RTX 3060). Should I stick with the 3060 or go for the 3060 Ti for some extra longevity?
3 Answers
If you're planning to run AI models locally, an 8GB card will really struggle and limit what you can do. I recommend aiming for at least 16GB for decent performance. If budget is tight, though, the RTX 3060 with 12GB could still get you started. Just keep in mind that for more advanced models, you might hit a wall.
It really comes down to what projects you'll undertake. The RTX 3060 should offer enough flexibility for beginners, plus you could even explore cloud options if needed. It lets you experiment with different workflows without overspending on a 3060 Ti right away.
I like that idea! A starting card sounds perfect for exploring my options.
Honestly, the more VRAM, the better. If you can grab the 3060 Ti, it might be worth it for a few extra bucks, but if you're just getting started, the 3060 will work for now. Just remember, for more complex tasks—like auto-coding—you’ll want more VRAM down the line.
Yeah, I’m just looking for a solid starting point for now. I can always upgrade later if I need more power.
That’s what I figured. Looks like I’ll have to settle for the 3060 since the budget for cloud options is also tight.