Green Benchmarking of BEAM Languages
summary
Built a Docker-based framework that measures energy (joules), throughput, latency and CPU/memory for Erlang, Elixir and Gleam web stacks under static, dynamic and WebSocket workloads. Results show a clear 30,000-request breakpoint where raw socket handlers stay in a low-energy band while framework stacks degrade; at 80k requests Elixir pure finishes ~64–66s and achieves highest throughput, whereas Gleam (mist) uses far more energy and fails on large WebSocket frames. Limitations: single laptop, colocated clients, frame-API differences.
AI-written summary · about
chapters (12)
- 0:00Green computing and BEAM
- 1:48Energy efficiency in programming
- 3:13Energy in web servers
- 4:58Erlang vs Elixir vs Gleam
- 7:09Benchmarking framework
- 12:39Visualizing results
- 13:00Static HTTP energy results
- 15:32Dynamic HTTP energy costs
- 16:35WebSocket energy use
- 19:10Study limitations
- 21:01Key takeaways
- 22:40Q&A
related talks
29mElixir - Evaluating Performance of Language
46mReal-world HTTP benchmarking
50mEnergy Efficient Design Principles of Software Engineering
54mReal-World HTTP Performance Benchmarking, Lessons Learned
30mAbove the Clouds: Investigating Systems from 32,000 ft.
32mSolid Ground
39mCreate an Eco-Friendly World with Green Software Engineering
31mCreate an eco-friendly world with Green Software Engineering