A tech enthusiast's ambitious homelab rebuild has ignited passionate discussions about the balance between computing power and practical necessity. The project, featuring eight mini PCs and a total investment of $13,500 USD, represents a significant leap from a simple two-machine setup to a full-scale Kubernetes cluster designed for serious infrastructure experimentation.
Hardware Specifications Comparison
| Component | Old Setup | New Setup |
|---|---|---|
| Compute | Mac Mini M1 (16GB), Beelink SER3 (16GB) | 8x Beelink SER5 Pro (32GB each), 1x Control Plane, 1x Database Server |
| Storage | Synology DS819 NAS | UniFi UNAS PRO with 7x 8TB Seagate Ironwolf |
| Networking | UniFi Dream Machine SE, UAP AC Pro | UniFi Dream Machine SE, U7 Pro, Pro HD 24 PoE Switch, 10G networking |
| Power | 1x UPS, 1x PDU | 2x UPS, 2x PDU |
| Rack Size | 9U Navepoint | 22U with sound treatment |
| Total Cost | ~$3,000-4,000 USD | ~$13,500 USD |
The Great Compute Debate
The community response reveals a fundamental divide in homelab philosophy. Many experienced users question whether such massive computing power serves any practical purpose for home use. Critics argue that the same learning objectives could be achieved with a single powerful machine running multiple virtual machines, potentially saving thousands of dollars while delivering superior performance per dollar spent.
The setup's compute-heavy approach has drawn particular scrutiny. Eight Beelink SER5 Pro units, each equipped with 32GB RAM and integrated GPUs, provide substantial parallel processing capability but raise questions about efficiency. Some community members suggest that consolidating the budget into fewer, more powerful machines with high-end GPUs would deliver better results for AI workloads and other demanding tasks.
Energy Costs and Environmental Concerns
Power consumption has emerged as a major talking point among homelab enthusiasts. While the new setup currently draws less than 200W at idle, community members share their own experiences with energy-conscious builds. Several users have downsized their labs specifically due to rising electricity costs, opting for ultra-low-power solutions like Intel N100 mini PCs that can handle substantial workloads while consuming minimal energy.
The contrast is striking when compared to alternative approaches. Some users report running comprehensive homelabs on single machines drawing 80W total, including NAS functionality and multiple services. This efficiency-first mindset reflects growing awareness of operational costs and environmental impact in personal computing projects.
Power Consumption Analysis
- Current Setup: <200W idle, up to 1,800W maximum
- Individual SER9 Units: 5-7W idle, 50-60W sustained load
- Monthly Cost: $12-20 USD (Idaho Power rates)
- Community Alternatives: 80-150W for complete homelab setups
- Energy-Efficient Options: Intel N100 systems at 10-15W total consumption
Storage and Reliability Concerns
The choice of UniFi's UNAS PRO for storage has generated significant criticism from the community. Users cite Ubiquiti's recent track record of software quality issues, including major bugs and problematic updates that have affected their networking products. Many recommend proven alternatives like TrueNAS or traditional Linux-based solutions for critical data storage, viewing reliability as non-negotiable for NAS applications.
This is definitely one of the purchasing decisions that I regret. My backups are robust and trustworthy enough that I don't have data loss concerns, but the software is atrocious and the customizability is extremely limited.
The Learning vs. Efficiency Trade-off
Despite the criticism, supporters argue that multiple physical machines provide unique learning opportunities that virtualization cannot fully replicate. Having separate hardware allows for experimentation with different operating systems, kernel modifications, and infrastructure components without risking the entire lab environment. This hands-on approach to distributed systems offers educational value that extends beyond what cloud services or single-machine setups can provide.
The project's focus on modern infrastructure tools like Kubernetes, GitOps, and declarative configuration management aligns with current industry practices. For professionals seeking to develop skills in these areas, having dedicated hardware for experimentation can justify the additional complexity and cost.
Software Stack and Tools
Infrastructure
- Kubernetes cluster running on Talos Linux
- PXE boot for easy upgrades and recovery
- Argo CD for GitOps deployment
- Democratic-csi for storage integration
Services and Applications
- n8n automation platform
- Home Assistant for smart home control
- Pocket ID for OIDC authentication
- Cloudflare Tunnels for external access
- GPU-accelerated AI workloads support
Conclusion
This homelab project highlights the ongoing tension between maximalist and minimalist approaches to personal infrastructure. While the $13,500 investment may seem excessive for home use, it reflects a growing trend of individuals building production-grade environments for learning and experimentation. The community's mixed response underscores that there's no one-size-fits-all solution – the right homelab depends entirely on individual goals, budget constraints, and tolerance for complexity. As energy costs rise and technology evolves, finding the sweet spot between capability and efficiency remains an ongoing challenge for enthusiasts worldwide.
Reference: Leveling Up My Homelab