SlowSort: An Enhanced Sorting Algorithm for Large‐Scale Integer Datasets

Background The performance of large‐scale integer deduplication sorting are critical, particularly in single‐threaded or resource‐constrained environments. Aims This study introduces SlowSort, an optimized algorithm based on the classic BitSort, aiming to enhance deduplication and sorting efficiency for integers, including negative values, while minimizing memory overhead. Materials & Methods SlowSort employs offset mapping to handle negative integers and uses the second smallest and largest values to define the bit vector length, excluding outlier effects, and It significantly improves single‐threaded execution efficiency through optimizations such as bitwise operations. Comparative experiments were conducted on small (10 7 range), medium (10 8 range), and large (2 32 range) integer datasets, benchmarking SlowSort against parallel sorting, radix sort, and standard library algorithms (Java Arrays.sort, C++ std::sort). Results On small datasets, SlowSort outperforms parallel sorting and matches radix sort. For medium datasets, its performance is comparable to both. On large datasets, SlowSort is 50% slower than parallel and radix sorts but reduces peak memory usage by 50%. Across all scales, SlowSort is over eight times faster than standard library algorithms. Discussion & Conclusion SlowSort offers an effective solution for large‐scale integer deduplication sorting in resource‐constrained environments, balancing performance and engineering practicality.