<![CDATA[Cryptographic hash functions are fundamental to transaction integrity in distributed-ledger systems, but their computational performance can vary across software and hardware environments. This study compares SHA-256, SHA3-256, and BLAKE2s-256 under a controlled single-node benchmark using identical simulated digital transactions. Transactions were serialized as canonical UTF-8 JSON and evaluated in batches of 50,000, 100,000, and 250,000 records. Each condition used five warm-up runs and 30 measured paired repetitions, producing 270 measured runs. Execution time, processing rate, descriptive statistics, 95% bootstrap confidence intervals, Friedman tests, and Holm-adjusted Wilcoxon signed-rank comparisons were analyzed. SHA-256 consistently produced the lowest mean execution time: 60.30 ms, 119.98 ms, and 290.88 ms for the three batch sizes, respectively. BLAKE2s required 94.42–441.39 ms and SHA3-256 required 97.97–470.14 ms. Differences among algorithms were significant at every batch size (Friedman p < 0.001; Kendall’s W = 0.903–0.968). These results indicate that SHA-256 was the most computationally efficient option in the tested Python/Windows AMD64 environment. The findings are implementation-specific and should not be interpreted as a universal security ranking of the algorithms.]]>
Copyrights © 2026
