Thank you to everyone involved in this project and thank you lwn for the article. I am a beginner in this space but my hobby is paralellism, coroutines, async

Thank you to everyone involved in this project and thank you lwn for the article. I am a beginner in this space but my hobby is paralellism, coroutines, async and multithreading. I read about the subinterpreter approach and I think the subinterpreter approach causes me to think of actors. You destroy the reference in the source interpreter when you send data to the other interpreter, like message passing. But it should be a O(1) message passing and transfer of ref counting responsibilities. The interpreters would be separate so you wouldn't have to worry about a dictionary being updated in two threads because they never touch the same objects. After reading https://blog.redplanetlabs.com/2023/08/15/how-we-reduced-the-cost-of-building-twitter-at-twitter-scale-by-100x/ https://blog.redplanetlabs.com/2023/08/15/how-we-reduced-the... I am thinking about how to parallelise the creation of relationships. What's a good practice for parallelising behaviour? If I want to fan out a message to 20 million collections, how can I use paralellisation to make it faster?