diff --git a/300 areas/360 Dev-Ops Network Computers/Home Network Map Overview.md b/300 areas/360 Dev-Ops Network Computers/Home Network Map Overview.md
index 7afe59f..e6966e6 100644
--- a/300 areas/360 Dev-Ops Network Computers/Home Network Map Overview.md	
+++ b/300 areas/360 Dev-Ops Network Computers/Home Network Map Overview.md	
@@ -54,9 +54,21 @@ aliases: []
 | **OS** | NixOS |
 | **SSH** | ✅ `sam@192.168.20.13` |
 | **Tailscale** | ✅ `100.114.62.46` (nixos-desktop) |
-| **Role** | Docker host, OmniRoute LLM proxy, always-on server. 15.5GB RAM.
+| **Role** | Docker host, OmniRoute LLM proxy, always-on server. 15.5GB RAM. GPU: GTX 760 (dead — see below).
 | **Key services** | Open WebUI v0.11.0 `:3000` (NixOS native, not Docker) |
 
+#### GPU upgrade — dead GTX 760 → AMD RX 6600 (recommended)
+
+**Status (2026-08-25):** GTX 760 (Kepler, PCI `10de:11c2`) no longer works. `nvidiaPackages.stable` dropped Kepler support after driver branch 470; dmesg shows `NVRM: does not include the required GPU ... probe failed (-1)`. The legacy 470 driver is EOL and very unlikely to build on kernel 6.18 (~10–25% odds even with kernel pinning).
+
+**Recommended replacement: AMD RX 6600** (~$180–210 USD)
+- **Zero NixOS driver pain:** in-kernel `amdgpu` — config becomes just `services.xserver.videoDrivers = [ "amdgpu" ];`. No legacy branches, no kernel-version roulette, ever.
+- **No PSU gamble:** many models need no PCIe power connector (132 W). Safe with the new PSU regardless of wattage headroom.
+- ~6–8× faster than the GTX 760; pairs cleanly with the Ryzen 5 5600. Flawless KDE/Wayland support.
+- Alternatives if a spare 8-pin + ≥450 W PSU confirmed: RX 7600 (~$250), RTX 3060 12GB (~$270, only if CUDA needed).
+
+**Swap procedure (safe):** power off → swap card → boot. If display fails, SSH recovery path is proven (`sam@192.168.20.27 → .13`); previous NixOS generations remain bootable from the boot menu. Update `videoDrivers` to `amdgpu`, then `nixos-rebuild switch`.
+
 | Docker containers | |
 |---|---|
 | **OmniRoute** | LLM proxy with combo routing (port 20128/20129). All pi LLM traffic routes through it. Combo `default-opencode-go-ds-flash` prioritises OpenCode-Go before falling through OpenRouter → DeepSeek direct → Google Gemini. See [[Pi Agent Extensions & Skills#OmniRoute — LLM Provider Router]] |