Comment by fdej

Yes, the worst-case complexity is exponential in N, but the wording in the article could lead you to believe that no explicit exponential bound is known, which is false.