Nathan Reynolds

When I set up my first standing desk next to a server rack, I underestimated how much the two systems would fight each other for floor space, cable runs, and thermal headroom. After two months of reorganizing, testing different motorized desks, swapping power strips, and rerouting CAT6 cables around a 12U half-rack, I finally built … Read more

I cut the cord back in 2018, and I have been running an HDHomeRun tuner ever since. The promise is simple: plug an antenna into a small box, run an Ethernet cable to your router, and stream free over-the-air broadcast TV to every screen in your house. After eight years of using these devices in … Read more

I tested the best mini PCs for 4K media playback over three months, running Plex transcodes, Netflix 4K streams, and Kodi on each one. My living room setup currently uses one of these as a dedicated home theater PC, and I’ve burned through enough HDMI cables to know what actually works in . The shift … Read more

I built my first home monitoring dashboard three years ago with a cheap Android tablet stuck to the wall with command strips. It worked for about six months before the battery swelled and the screen started ghosting. That experience taught me that picking the right hardware matters just as much as the software you run … Read more

I have been running a self-hosted Jellyfin server for over three years now, and during that time I have tested more than 20 different Android TV boxes on my living room TV. Some worked great. Some forced transcoding on every 4K remux. A few even crashed mid-movie when bitstream audio kicked in. That is why … Read more

I have walked into enough messy server rooms to know that a missing or peeling label is a real headache. After testing 8 label makers over six weeks in a live data center environment, I can tell you the Brother P-touch vs NIIMBOT for server labels decision is not as simple as picking whichever is … Read more

I spent the last three months running 70B parameter models on six different compact workstations, and the Ryzen AI Mini PC vs Mac Studio debate is finally getting interesting. When the Strix Halo generation launched, Apple Silicon had the local LLM scene mostly to itself. Now AMD’s Ryzen AI Max+ 395 ships with 128GB of … Read more

I spent the last three months running Stable Diffusion locally on eight different GPUs to find out which ones actually deliver on the promise of fast, high-quality AI image generation. I generated thousands of images, trained LoRAs, pushed ControlNets to their limits, and watched VRAM meters like a hawk. What I found changed how I … Read more

I spent the last three months running GPU passthrough benchmarks across our Proxmox test cluster, swapping eight different graphics cards into a Ryzen 9 7950X rig with 64GB of DDR5 memory. After 47 test runs covering gaming VMs, LLM inference, and Plex transcoding workloads, I can confidently say the gap between “technically works” and “actually … Read more

Your Home Assistant hub is the brain of your smart home, but the moment the grid drops, that brain goes dark. I learned this the hard way when a storm knocked out power for 14 hours and my entire Zigbee mesh, Z-Wave locks, and Home Assistant automations went silent right alongside the fridge. A reliable … Read more