Three and a Half Weeks, 689M Tokens, No Meter
689M tokens. 25 days. ~$40 electricity. The math, the caveats, and what changed when the meter stopped existing.
Qwen3.8 27B on Dual RTX 3090s: A vLLM Field Report
Three days into making Qwen3.8 27B my daily local coding driver on dual 3090s — the hybrid linear-attention architecture, an OOM that masqueraded as a backend error, and what the …
Highly Available AI Inference Cluster
What this proves: Resilience — enterprise-grade AI infrastructure at a single operator's desk. A custom routing plane in front of a three-server GPU fleet that switches models and …
•
3 min read
Read more
Running Qwen3.6 27B Locally on Dual RTX 3090s with vLLM v0.19
How I went from a blank Docker template to 116+ tok/s with speculative decoding, FlashInfer, and a 160k context window on dual 3090s.
The Proxmox Utility Toolkit: Stop Cloning That VM by Hand
A tour of my open-source Proxmox utility toolkit — scripts, docs, and learning paths for homelabbers and enterprise engineers who are tired of repeating themselves.
Nagios: A 10-Year Retrospective on Infrastructure Monitoring
A decade-long retrospective on implementing Nagios for enterprise infrastructure monitoring, from homelab validation to production deployment at scale.
Self-Hosted AI Inference Cluster
I build and operate AI infrastructure right now, today.
•
1 min read
Read more
Upgrading from RHEL 7 to RHEL 8
Insights on upgrading from rhel 7 to rhel 8.
Comparing Docker.io and Snap Docker on Ubuntu
Key differences between installing docker.io and snap docker on Ubuntu include file access permissions, affecting security and flexibility
Self-Hosted Privacy-First Analytics
Plausible Analytics deployed as a self-hosted alternative to Google Analytics in Podman containers behind Cloudflare tunnels.
•
1 min read
Read more
Showing 1-10 of 13 items (Page 1 of 2)