I'm working on a new PC setup aimed at machine learning, coding, general workspace tasks, and using Ableton. I want to make sure my part list works well together, especially focusing on finding the smallest case possible without sacrificing thermal performance too much (ideally under 5%). Here's what I'm considering so far for my build:
**Case Options:** AP201, NCASE N2, Cerberus X, or Lian Li A3
**Specs:**
- **CPU:** AMD Ryzen 9950X
- **GPU:** Nvidia 3090
- **RAM:** Corsair Vengeance EXPO/XMP DDR5 6000MHz (CL36)
- **Motherboard:** ASUS TUF Gaming B850M-PLUS WIFI
- **SSD:** Crucial T500 2TB NVMe Gen4 with Heatsink
- **Power Supply:** ASUS TUF 1000W Gold ATX3.0
- **CPU Cooling:** Arctic Liquid Freezer II 360 AIO
- **Fans:** Arctic P14/P12 wherever necessary
- **Extras:** Anti-sag GPU bracket
I'd love your thoughts on whether this setup will work well and if my case choices are sound!
2 Answers
Interesting choice going with a two-gen-old GPU like the 3090. Is there a reason behind that?
I'm building in the Lian Li A3 soon. It’s compact and fits more than you’d expect. Just make sure to include a few case fans for airflow. Definitely helps to have more intake than exhaust for positive pressure. But yes, smaller cases will naturally run a bit hotter than mid or full towers, but it should still be manageable!
Right, but the reviews for the Lian Li A3 seem mixed compared to the AP201. I'm leaning towards the AP201 myself mainly because it seems to have a better reputation.
You think the temperatures would be much worse in super compact cases like the NCASE N2?
Definitely! The 3090 still packs a punch for ML with its 24GB of VRAM. Plus, the next-gen cards jumped from 16GB on the 5080 to 32GB on the 5090, so the 3090 can actually be a better option for some tasks.