2000 lines is nothing.
The performance differences between one large file and many smaller ones are negligable.
The main reason for splitting things up is code-reuse.
For example, if you have a set of functions that talk to your database, you may want to use the same functions in another script too, so you'd put them into a seperate file. That way if you update the functions (add new functions, remove bugs) the other scripts automatically use the new versions too.
A few general notes:
Long scripts take longer to parse.
Many smaller files take longer to load, and they still have to be parsed.
If your total number of lines does not change between the large file and the many smaller ones, then using one large file is probably faster, but also more difficult to maintain.