johnc602

Members
  • Posts

    1
  • Joined

  • Last visited

johnc602's Achievements

Noob

Noob (1/14)

0

Reputation

  1. unRAID 6.11.0 - Plus License Asus TUF Z390 MB intel i7-8700 32 GB GSkill Ripjaws RAM Just got unRAID up and running a couple of weeks ago and loving it so far. Previously I had a Synology and a Proxmox server and I'm liking the power/flexibility of unRAID a lot better. My setup is an ASUS TUF Z390 with an i7-8700, 32GB of RAM and four spinning disks. I'm hoping to add two video cards for VM passthrough, a 10G card and two NVMe drives for cache. Existing Persistent Storage: 3-4 spinning disks (no issue here) To be added Cache: 2 NVMe drives (likely 1 TB ea). Thoughts are to use either both in RAID1 or separates for docker and VM storage To be added NIC: 10G (thinking Mellanox Connectx3 10G) To be added GPUs: 1 for passthrough for windows 10VM for light gaming (Nvidia RTX3XXX) and 1 for passthrough to a Mac VM (AMD Radeon) Questions: - My motherboard only has 2 PCIe x16 slots and a few x1 slots. Is it possible to run both GPUs (x16 or x8?) AND the Mellanox card (requires x4 as I understand it)? Are there any drawbacks to performance on any of these in this config? If I can only run one GPU and the Mellanox card, does it make sense to try to pass the iGPU through to the Mac VM? - Can I run both NVMe drives without impacting the performance of the GPUs? Any implications to SATA performance (outside of losing 2 of the 6 SATA ports on the board)? - How do I handle write caching to the array for things like downloads if my two NVMe drives are tied up with docker and VMs? Sorry for the basic questions, I haven't pushed the limits of PCIe like this before and don't fully understand the lanes/bandwidth and the disabling of peripherals when others are plugged in (e.g. I think both NVMe drives plugged in will disable 2 SATA ports) I might be trying to do too much with one box here, but my hope was since most of these different services/apps will not be used simultaneously I could leverage a single MB/CPU/RAM across all.