Sudoku can be regarded as a SAT problem. Various encodings are known for encoding Sudoku as a Conjunctive Normal Form (CNF) formula, which is the standard input for most SAT solvers. Using these encodings for large Sudoku, however, generates too many clauses, which impede the performance of state-of-the-art SAT solvers. This paper presents an optimized CNF encodings of Sudoku to deal with large instances of the puzzle. We use fixed cells in Sudoku to remove redundant clauses during the encoding phase. This results in reducing the number of clauses and a significant speedup in the SAT solving time.