Hi @nxp
I’ve been testing an environment where NXP’s PEX hardware is used as part of a backend cluster that supports a virtual desktop infrastructure (VDI). The idea is to connect thin clients to a central host through VDI, but I’m running into a few challenges.
Main things I’m trying to figure out:
How to minimize latency and jitter when devices (USB, storage, etc.) run through multiple layers (PEX → hypervisor → VDI).
Whether GPU passthrough or SR-IOV is stable enough in this kind of setup.
Any best practices for isolating devices securely between different VMs.
So far, I’ve tried pinning some devices to dedicated PCIe lanes and doing a basic GPU passthrough test, but results are mixed. Before I dive deeper, I wanted to ask:
Has anyone here combined NXP PEX hardware with a VDI environment? If yes, what issues did you face, and what worked best for you?
Appreciate any insights or experiences.
