In what way do major chess engines differ from one another?
Do they all use the same algorithm and the same evaluation function, differing only in parameters/weights?
Lot's of possibilities (I can't list everything):
While all chess engines I'm aware of use alpha-beta/nega-max, their implementation could be significantly different.
No. They don't all use the same algorithm and the same evaluation function. Visit https://chessprogramming.wikispaces.com/Home, and you'll appreciate the complexity in chess engine programming.