The aim of this paper is to examine and demonstrate how integer-based datetime labels (integer surrogate keys for time) can optimize data-warehouse and time-series performance, proposing practical formats and algorithms and validating their efficiency on real-world workloads. It is shown that replacing standard DATE and TIMESTAMP types with 32- and 64-bit integer formats reduces storage requirements by 30-60 percent and speeds up query execution by 25-40 percent. The paper presents indexing, aggregation, compression, and batching algorithms demonstrating up to an eightfold increase in throughput. Practical examples from finance, telecommunications, IoT, and scientific research confirm the efficiency and versatility of the proposed approach.

翻译：本文旨在探讨并论证基于整数的日期时间标签（时间的整数代理键）如何优化数据仓库与时间序列性能，提出实用的格式与算法，并在实际工作负载中验证其效率。研究表明，用32位与64位整数格式替代标准DATE与TIMESTAMP类型，可降低存储需求30-60%，并加速查询执行25-40%。本文提出的索引、聚合、压缩与批处理算法，展示了高达八倍的吞吐量提升。来自金融、电信、物联网及科学研究的实际案例，证实了所提方法的高效性与普适性。