Community Forums

Hey everyone!

I've finally completed my new workstation build, optimized for demanding tasks like high-poly 3D rendering in Blender and 4K video editing in Premiere Pro. I wanted to share the specs and some insights for anyone looking to build a similar machine.

CPU: AMD Ryzen 9 7950X3D
Cooler: Noctua NH-D15
Motherboard: ASUS ROG STRIX X670E-A GAMING WIFI
RAM: 64GB (2x32GB) DDR5-6000 CL30 Corsair Vengeance
GPU: NVIDIA GeForce RTX 4090 24GB GDDR6X
Storage:

• 1TB Samsung 980 Pro NVMe SSD (OS/Apps)
• 2TB Samsung 990 Pro NVMe SSD (Projects)
• 4TB WD Black HDD (Archive)

PSU: Corsair RM1000e (1000W 80+ Gold)
Case: Lian Li Lancool III RGB
OS: Windows 11 Pro

The build was relatively straightforward, but cable management in this case took some time. Performance so far has been incredible. Rendering times have been slashed significantly compared to my old setup. Let me know if you have any questions about specific components or the build process!

Wow, that's a beast of a machine, TechGuru88! The 7950X3D and 4090 combo is exactly what I'm aiming for. How are your temps under heavy load with the NH-D15? I was considering an AIO, but the NH-D15 is tempting.

Also, did you encounter any issues with the X670E board and the 7950X3D in terms of BIOS compatibility or general stability? I've heard some mixed reviews about early AM5 platforms.

Hey PCWiz!

Thanks! Temps have been surprisingly good. Under full load (Cinebench R23 multi-core), the 7950X3D peaks around 85-88°C, and the 4090 stays in the low 70s. The NH-D15 is a phenomenal air cooler, and I haven't felt the need for an AIO. Plus, less complexity!

Regarding the X670E board, I updated the BIOS to the latest version (version 0703 as of my build) right out of the box, and it's been rock solid. No stability issues or weird behavior. The 7950X3D runs beautifully with it. Make sure to grab the latest BIOS from ASUS's support page for your specific board model.

Great build! I'm in the process of upgrading my own workstation. How does the 4090 handle Houdini simulations compared to CUDA-accelerated tasks?