To get started, check out the new Raspberry Pi Pico section in the <a href="https://github.com/eclipse-zenoh/zenoh-pico?tab=readme-ov-file#226-raspberry-pi-pico">README</a>. This includes detailed instructions for building and running Zenoh-Pico on Pico devices, enabling seamless integration into your IoT projects.</p><h3 id=performance>Performance</h3><p>A lot of effort has been devoted to improving throughput and latency performance in Zenoh-Pico. A big improvement has been brought by:</p><ul><li>The rework of the internal fragmentation engine for large messages.</li><li>The addition of manual batching for small (non-fragmented) messages.</li></ul><p>Notably, the latency has been reduced to up to 1/50th (i.e. 5000% better) for large fragmented messages and up to 35% for non-fragmented messages. Moreover, the throughput has increased up to 60% for non-fragmented messages and up to 10x (i.e. 1000% better) for large fragmented messages in peer-to-peer over UDP multicast (§). In client mode, the throughput has been increased up to 10x (i.e. 1000% better) for large messages.
0 commit comments