Microsoft this morning released a test version of Excel 2010 that the company says will be able to tap into distributed clusters of computer servers to crunch huge volumes of data much more quickly than before. You aren't going to need this to complete your weekly expense report -- well, hopefully not -- but it could come in handy for scientists and others that work with very large Excel workbooks.
The company made the announcement at the SC09 supercomputing conference in Portland. The functionality works in conjunction with Microsoft's Windows HPC (High-Performance Computing) Server 2008 R2, which was released in beta form this morning and is due out in final form next year.
To give a sense for the way it juices up Excel, the company cites the example of a life-insurance actuarial firm that was able to reduce the amount of time to process 1,700 records from 14 hours to less than three minutes. Processing 1 million records previously took more than seven days, but was reduced to about two hours.
The feature in Windows Server HPC works by running distributed instances of Excel on multiple servers, said Vince Mendillo, Microsoft's senior director of High Performance Computing. That means workbooks can be calculated in parallel across the cluster, he explained. Microsoft has said the next versions of Excel and other Office 2010 programs will be released in the first half of next year.
Seattle-based supercomputer company Cray this morning separately announced a new integrated workstation that uses Windows 7 and Windows HPC Server 2008.
No comments:
Post a Comment