Distributed Graph-Based Task Executions and Computation Caching

Nx is smart. It analyzes your workspace and figures out what can be affected by every code change. That's why Nx doesn't rebuild and retest everything on every commit — it only rebuilds what is necessary.

Nx partitions commands into a graph of smaller tasks. Nx then runs those tasks in parallel, andit can even distribute them across many machines without any configuration.

Nx also uses a distributed computation cache. If someone has already built or tested similar code, Nx will use their results to speed up the command for everyone else.