How to Run DeepSeek AI Models with NVIDIA GPU Passthrough in Proxmox

Running AI models locally provides enhanced privacy, reduced latency, and complete control over your infrastructure. In this guide, we’ll walk through setting up DeepSeek models (like DeepSeek-R1) inside a Proxmox virtual machine with GPU passthrough. This configuration allows the AI model to access your NVIDIA GPU directly for significantly improved performance.

Podman 5.5.0 released

Version
5.5.0
of the Podman container-management tool has been
released. Notable features include the addition of a podman machine cp command to copy files into a running Podman
VM
, a podman artifact extract command to copy
contents of an OCI
artifact
to disk, and a --mount=artifact option to mount
OCI artifacts into containers. See the release announcement for a full
list of improvements and bug fixes.

[$] Faster firewalls with bpfilter

From
servers in a data center to desktop computers, many devices
communicating on a network will eventually have to filter network
traffic, whether it’s for security or performance reasons. As a result,
this is a domain where a lot of work is put into improving performance:
a tiny performance improvement can have considerable gains.

Bpfilter
is a
project that allows for packet filtering to easily be done with BPF, which can
be faster than other mechanisms.

AMD Ryzen 9 9900 Series Linux Performance Since Launch

After recently looking at how the Intel Core Ultra 9 285K “Arrow Lake” Linux performance has evolved since launch, many Phoronix readers were curious how a similar launch-day vs. now comparison would look on the AMD Zen 5 side. The article today is looking at how the AMD Ryzen 9 9900X and Ryzen 9 9950X Linux performance has evolved since their launch last year. These numbers are put alongside the prior Intel Arrow Lake results for additional context.