Performance Evaluation and Acceleration for XML Data Parsing

As XML has become a standard for data representation and exchange, XML data processing becomes more and more important for server workloads like web servers and database servers. One of the most time consuming part is data parsing. This paper presents a performance study of XML data parsing through measurement and simulation. They evaluate the data parsing behavior and study architectural characteristics such as instruction mix and cache behavior. It is observed that XML data parsing benefits from a larger instruction cache compared to data cache. Based on the analysis, they incorporate new instructions with special hardware support to speedup certain frequently-used operations.