Which Programming Languages Are the Most Energy Efficient? Insights from a 2017 Study

When a phone is low on battery, its management software often prompts us to close high‑consumption apps. This shows that, besides hardware manufacturers, software developers also need to pay attention to energy usage.

In this article, researchers examine the energy consumption of various programming languages.

In 2017, a team of six Portuguese researchers published the paper “Energy Efficiency Across Programming Languages”. They implemented ten benchmark problems in 27 languages using the same algorithms, then measured each language’s power consumption, execution speed, and memory usage.

Their findings include:

Fast execution does not necessarily mean low energy consumption.

Compiled, virtual‑machine, and interpreted languages were compared, as were imperative, functional, object‑oriented, and scripting paradigms.

The benchmarks were based on the Computer Language Benchmarks Game, ensuring comparability.

Overall, C was the fastest and most energy‑efficient, but in specific tasks (e.g., DNA‑sequence scanning) Rust consumed the least power while C ranked third.

The “best” language varies with the metric considered; a language that ranks high in speed may rank lower in memory usage or energy.

Measurements were taken on a Linux Ubuntu Server 16.10 desktop (kernel 4.8.0‑22‑generic) using Intel’s Running Average Power Limit tool, running each program ten times to reduce cold‑start and cache effects.

Is Faster Always Greener?

The study challenges the common assumption that faster programs consume less energy. Power draw is not constant over time, so a program that finishes quickly may still use more energy than a slower one, as shown by a Chapel vs. Pascal comparison.

Overall, about 88 % of the energy consumption comes from the CPU, regardless of whether the language is compiled, interpreted, or runs on a virtual machine.

Compiled Languages Lead in Efficiency

On average, compiled languages required 120 J of energy, versus 576 J for virtual‑machine languages and 2 365 J for interpreted languages. Execution times averaged 5 103 ms for compiled, 20 623 ms for VM, and 87 614 ms for interpreted languages.

The five slowest languages were all interpreted: Lua, Python, Perl, Ruby, and TypeScript. The five most energy‑intensive languages were also interpreted: Perl, Python, Ruby, JRuby, and Lua.

However, in regex‑heavy string processing, three interpreted languages (TypeScript, JavaScript, PHP) were among the most efficient.

In memory usage, compiled languages also ranked among the lowest, with average footprints of 125 MB (compiled), 285 MB (VM), and 426 MB (interpreted). When grouped by paradigm, imperative languages used the least memory (≈116 MB), followed by object‑oriented (≈249 MB), functional (≈251 MB), and scripting (≈421 MB).

Choosing the Right Language

The optimal language depends on the specific constraints of the application. For tasks requiring both low energy and fast execution, C is often the best choice. If minimizing memory is also a priority, C, Pascal, and Go are good candidates. Different paradigms and implementations can lead to varied results.

Conclusion

Software development is moving toward greener programming and modernizing legacy systems to meet 21st‑century standards. Are you ready to adopt energy‑aware coding practices?