The Hardware#
I didn’t buy anything new. This is the machine I had:
| Component | Spec |
|---|---|
| CPU | Intel i5-9600K — 6 cores / 6 threads @ 3.7GHz |
| RAM | 64GB DDR4 |
| GPU 1 | RTX 2080 Ti — 11GB VRAM (CUDA compute) |
| GPU 2 | GTX 1060 3GB (display + overflow) |
| OS | Ubuntu 22.04, CUDA 12.5 |
| Storage | NVMe SSD |
The RTX 2080 Ti is the key. 11GB of VRAM is enough to run serious local models, Blender GPU rendering with OptiX, and Kokoro TTS simultaneously.
What’s Running On It#
graph TD
GW["🦞 OpenClaw Gateway
(port 18789)"]
NEO["🧠 Neo4j KG
(port 7687, Docker)"]
CHR["🔍 ChromaDB
(vector memory)"]
PLEX["🎬 Plex + *arr stack
(ports 32400, 7878, 8989...)"]
DOCKER["🐳 Docker (Colima)"]
GPU1["⚡ RTX 2080 Ti
Blender OptiX
Kokoro TTS
Ollama models"]
XVFB["🖥️ Chrome on Xvfb:99
(X autopilot, CDP:9222)"]
NETDATA["📊 Netdata v2.8.5
(monitoring)"]
GW --> NEO
GW --> CHR
DOCKER --> NEO
DOCKER --> PLEX
GPU1 --> GW
Always-on services:
- OpenClaw gateway — the AI brain (WebSocket, port 18789)
- Neo4j — knowledge graph memory (Docker)
- ChromaDB — vector memory
- Plex + qBittorrent + Radarr + Sonarr + Prowlarr
- Netdata monitoring dashboard
On-demand:
- Blender GPU rendering (OptiX on RTX 2080 Ti)
- Kokoro TTS (local voice synthesis, zero cost)
- Chrome on Xvfb + CDP for automation
- Ollama local models
Power & Cost Breakdown#
| Scenario | Power Draw | Monthly Cost |
|---|---|---|
| Idle (gateway + Docker) | ~100W | €25 |
| Active AI agents | ~150-200W | €35-40 |
| GPU rendering burst | ~500W | Peaks only, averages out |
| Realistic average | ~130W | ~€30/month |
Now compare to what I’d pay on AWS:
| Cloud Option | Monthly Cost | GPU? |
|---|---|---|
| t3.medium (4GB RAM) | ~$30 | ❌ |
| t3.large (8GB RAM) | ~$60 | ❌ |
| g4dn.xlarge (T4 GPU) | ~$380 | ✅ (16GB) |
| Gaming server (gaming PC I own) | €25-30 | ✅ (11GB VRAM) |
The gaming server wins on every axis except latency and uptime guarantees. For a dev/staging setup and personal AI infra, that’s fine.
Remote Access#
The machine sits in my home but I access it from everywhere via Tailscale:
# From my Mac
ssh amine@your-device.ts.net
# OpenClaw TUI
openclaw tui --url wss://your-device.ts.net --token <token>
Tailscale gives me a stable hostname regardless of ISP changes, with WireGuard encryption. The gateway binds to 127.0.0.1 and Tailscale Serve proxies it over HTTPS — no port forwarding, no exposed ports.
What Surprised Me#
The 64GB RAM matters more than the GPU. Running 5 concurrent Claude sub-agents means 5 × ~400MB Node.js processes plus Neo4j (800MB) plus Docker containers. 16GB would OOM immediately. 64GB means I never think about it.
No swap was a mistake. I was running with zero swap on Ubuntu until an agent storm caused an OOM kill. Added 2GB swap as a safety net — it’s never actually used but it prevents hard crashes.
The GTX 1060 is useless for CUDA. SM 6.1 isn’t supported by modern PyTorch. It handles display output and that’s it.
Is This Worth It?#
If you already own the hardware: absolutely yes. The break-even vs. renting a GPU cloud instance is immediate. I run:
- A production AI agent stack
- Local video rendering
- A media server
- All monitoring and backups
…for less than a t3.large with zero GPU. The tradeoff is uptime (home internet can hiccup) and maintenance time. For personal AI infrastructure, both are acceptable.
The real unlock was pairing it with Tailscale + OpenClaw. Now my phone and Mac both connect to the same AI brain running on this machine, and the brain can use the GPU.