An Evaluation of Real-Time Performance for Nerves

In this video, you’ll hear about the results of Hideki Takase’s research and learn how to judge the real-time performance of IoT systems built with Nerves.

This talk originally aired on October 29, 2020.

Real-time performance means the property of being able to finish processes up to a specified time. Erlang and Elixir are known to have soft real-time performance. So Hideki Takase was then curious about how much real-time performance Nerves has. The purpose of his research was to clarify the real-time performance of Nerves. Specifically, he evaluated the variation in execution time of Nerves applications caused by changing the heap size for GC, the usage of NIF, and priority settings of Erlang processes using different hardware.

In this video, you’ll hear about the results of his research and learn how to judge the real-time performance of IoT systems built with Nerves.