The PivotTable Service is the client side of Microsoft SQL Server OLAP Services. It provides the layer that lets client applications connect with SQL Server OLAP Server. You can use OLE DB for OLAP or ActiveX Data Objects Multi-Dimensional (ADO MD) to write code to access and manipulate data cubes. ADO MD lets you develop Web applications against OLAP Server cubes.

The PivotTable Service can cache user queries, meta data, and data so that it can satisfy new queries by working with previously cached data rather than requerying the OLAP Server. The OLAP Server and PivotTable Service share a great deal of code. This sharing provides multidimensional calculation, caching, and query management features to the client. For example, if you query the OLAP Server for first-quarter 1998 sales information and then decide to compare the results to first-quarter 1997 data, the PivotTable Service will query the server only for the 1997 data because the 1998 data is already cached. In addition, the PivotTable Service facilitates disconnected operations by allowing cube slices to persist on the client machine for later analysis, without requiring a connection to the OLAP Server.

According to Microsoft, the PivotTable Service uses approximately 2MB of disk space and 500KB of memory. Office 2000 will include the PivotTable Service, and the version of Excel in Office 2000 will include embedded connectivity between spreadsheets and OLAP Server data or other OLE DB for OLAP providers.