> So you’re stuck choosing a single CPU or having to write send and sync everywhere. There’s a lot of use cases where you would want a thread-per-core model lik

> So you’re stuck choosing a single CPU or having to write send and sync everywhere. There’s a lot of use cases where you would want a thread-per-core model like Glommio to take advantage of multiple cores while still being able to write code like it’s a single thread. No your not, you spawn a runtime on each thread and use spawn_local on each runtime. This is how actix-web works and it uses tokio under the hood. https://docs.rs/actix-rt/latest/actix_rt/ https://docs.rs/actix-rt/latest/actix_rt/