Questions about code splitting often come up in the context of optimizing frontend applications. This approach is used every day, so you need to understand it for more than just interviews.
Code splitting is a page load optimization technique where we split JavaScript code into several smaller parts (bundles) and, instead of loading one large file, load several small parts in parallel.
Besides parallel loading, code splitting lets us load code not all at once, but only when we actually need it. This way we can load code in the background so it doesn't affect UX.
For example, say we have one large JavaScript file used on an HTML page. Without code splitting, the browser has to spend more time loading it. If the user has a poor internet connection, you may accidentally drive them up the wall, and they won't wait for your site to load.
Now imagine this is a large CRM and most of that code isn't needed for the first load. That's exactly why code splitting is used. It lets you reduce the size of the initial JavaScript bundle, show the first content to the user faster, and load the additional code in the background or on demand.