5 August 2003 Oracle will next month take the wraps off a new version of its database system designed to exploit the distributed, high-performance characteristics of grid computing within enterprise environments.

At Oracle OpenWorld in San Francisco, California in early September, the company will unveil Oracle 10G, an implementation of the database designed to execute resource-intensive data warehouse queries across large networks of under-utilised servers.

Oracle claims that 10G cracks one of the major problems that has prevented the application of grid computing outside of scientific communities — the ability to prioritise the execution of jobs. While many large-scale scientific processing tasks can be executed when the dispersed resources become available, enterprise workloads require predictable capacity, the company says.

The system will enable organisations to make use of unused processing power distributed across their networks. Reports also suggest that 10G will also be able to handle massive, distributed databases, scaling to eight exabytes.

Pre-announcement details of 10G have been filtering out ever since a page on the company’s web site inadvertently highlighted a 10G product presentation scheduled for October in Paris.

Oracle partner Hewlett-Packard (HP) also let slip that a recent record-breaking data query benchmark used 10G. A cluster of HP servers running the new database implementation set a new transaction throughput high for the TPC-C benchmark.