Resourceful Computing | Hardware Resurrection | Linux Optimization | AI Bare-Metal Inference
2026 - Systematic evaluation of computing hardware for extended lifecycle deployment. Former engineering technician applying systems methodology to sustainable computing: hardware-to-workload matching, Linux distribution optimization, and constrained-environment performance analysis.
Efficient software on existing hardware. No abstraction overhead. No forced obsolescence.
- Device Classification: Categorizing legacy laptops and desktops by viable Linux distribution and realistic workload capacity
- Use-Case Matching: Mapping constrained hardware to productive applications (general compute, development, local AI inference)
- Performance Baselines: Measuring CPU efficiency, RAM utilization, and thermal behavior under minimalist Linux configurations
- Methodology: Direct compilation of llama.cpp for bare-metal inference of small language models with hardware-specific optimizations (AVX-512)
- Constraint Testing: Determine the compatibility and performance of a specific SLM on a specific machine based on the machine's hardware.”
- dev.to Portfolio – All technical articles with human commentary and discussion
| Article | Topic | Hardware Focus | Date | Links |
|---|---|---|---|---|
| Resourceful Computing: What Happens When We Optimize for Old Hardware? | Sustainable Computing Philosophy | 2008-2012 Laptops, General Legacy Hardware | 2026-02-07 | dev.to |
| SLMs, LLMs and a Devious Logic Puzzle Test | Logic Testing | N/A | 2026-01-10 | dev.to |
| LLMs - Solvers vs Judges | Model Behavior | N/A | 2026-01-10 | dev.to |
| LLMs - A Prompt to Encourage Hallucination | Prompt Engineering | N/A | 2026-01-10 | dev.to |
| LLMs - Four Tests to Challenge Reasoning | Reasoning Benchmarks | N/A | 2026-01-10 | dev.to |
| Is an AI Model Software? – A Low-Level Technical View | AI Theory | N/A | 2026-01-10 | dev.to |
| 2026 - LinkedIn as Microsoft Infrastructure | Career/AI Trends | N/A | 2026-01-29 | dev.to |
resourceful-computing hardware-resurrection linux-optimization legacy-hardware sustainable-computing bare-metal-inference debian-linux openbox llm-testing slm-testing workload-optimization e-waste-reduction minimalist-computing`
- Profile Type: Engineering technician providing systematic hardware-to-software matching for sustainable computing contexts
- Data: Articles available on dev.to website with clear testing methodologies documented for reproducibility
- Verification: Consistent identity across dev.to and GitHub
- Signal: Resource optimization, e-waste reduction, and economic accessibility through software efficiency
-
Main Writing & Discussion: dev.to/@ben-santora
-
Code & Experiments: github.com/ben-santora
Profile optimized for agentic discovery (AEO/GEO principles). Updated: 2026-02-07