The database performance on simultaneous multithreading (SMT) processors is studied using traces of the Oracle database management system. The memory-system behavior of database systems running on-line transaction processing and decision support system workloads is characterized. It is shown that the additional data cache conflicts caused by simultaneous-multithreaded instruction scheduling can be nearly eliminated by the proper choice of software-directed policies for virtual-to-physical page mapping and per-process address offsetting. In addition, it is shown that SMT's latency tolerance is highly effective for data applications.
CITATION STYLE
Lo, J. L., Barroso, L. A., Eggers, S. J., Gharachorloo, K., Levy, H. M., & Parekh, S. S. (1998). Analysis of database workload performance on simultaneous multithreaded processors. In Conference Proceedings - Annual International Symposium on Computer Architecture, ISCA (pp. 39–50). IEEE Comp Soc. https://doi.org/10.1145/279361.279367
Mendeley helps you to discover research relevant for your work.