Using DataSynapse GridServer technology, organisations can create a cost-effective High-Performance Computing (HPC) environment by tapping into the unused processing and storage capacity of existing servers across the enterprise. Applications that are virtualised can be split into functional components, such as individual calculations, modelling and simulations, that can all be processed in parallel across a distributed computing infrastructure, significantly reducing the overall processing time.
At the investment banking division of one of the UK's biggest global banking groups, for example, it was taking two hours each night to upload and aggregate 500,000 lines of financial transaction data from the bank's General Ledger system into the Longview database to support the bank's daily reporting requirement. By virtualising the Longview application using the DataSynapse GridServer technology so that billions of complex calculations could be performed in parallel by a large number of servers and desktops across the bank's global computing infrastructure, the processing time for complex calculations has been reduced by more than 90 percent.
"The bank had already invested in GridServer technology from DataSynapse to accelerate the performance of its trading and risk management applications," says Kevin Kelly, Director, Business Development, EMEA at DataSynapse. "The bank was able to simply and quickly add the Longview application to the existing computing grid, significantly improving their ability to perform timely, accurate financial reporting based on massive amounts of data - an otherwise difficult task for any performance management tool."
The Longview-DataSynapse solution addresses a challenge familiar to many companies - uploading financial data from a GL system, along with data from other systems such as HR and ERP, to a multi-dimensional performance management tool, where various calculations can be performed, comparisons with budgets and forecasts made, and reports generated on the performance of the business. For most companies using most performance management tools, this process takes place quarterly, monthly or perhaps weekly. In a GridServer environment, the Longview application can do this on a daily or more-frequent basis, even for the largest data requirements. This is especially helpful for large companies processing large volumes of data that want real-time reporting capabilities.
"By virtualising the Longview application with GridServer technology, we can now perform complex calculations on billions of rows of data in a fraction of the time," says Mark E. Burton, President and CEO, Longview Solutions. "This is great news for Longview customers, who can now leverage existing computing capacity worldwide to achieve dramatic new levels in performance and scalability, while still getting all the benefits of a single application and a single repository of data that have made the Longview Platform unique in the industry."