CPU near 100% -> packet loss?

[tortoisedoc]

Setup:

• 3 separate physical servers setup, connected via Wireguard on internal network
  Problem:
  We have noticed that on high load (during performance testing), the slirp4netns nears 100% CPU (expected).
  This however seems to cause packet loss, and consenquent timeouts at HTTP level (for example to kubernetes control node).
  The configuration is the one from usernetes (i.e. no new options added, neither old removed).
• is there ways to improve slirp4netns performance (i.e. by scaling out to more CPU cores)
• is there means for us to investigate more?