Hello, Excel 2007 already does parallel computing. A spreadsheet program internally creates a dependency graph of the cells. A dependency graph also knows which parts of a complex calculation are not dependent on each other, and therefore can be run in parallel. Excel 2007 does this parallel computing on different cores on multicore machines. Using different machines in a cluster is a technical change, but all the concepts are already implemented if you can do it on a multicore machine.
Kohei Yoshida wrote: > Hello Maho-san, > > On Thu, 2009-11-19 at 12:34 +0900, Maho NAKATA wrote: >> http://news.cnet.com/8301-13860_3-10398590-56.html >> >>> At a key supercomputing conference on Monday, Microsoft released >>> a test version of its Excel spreadsheet redesigned to run on powerful >>> clusters of servers. >> Very interesting news to me. I didn't attend SC09, though. >> >>> By engineering Excel to run better on such clusters Microsoft said that >>> customers are seeing spreadsheets that normally would take weeks to >>> calculate >>> now run in a few hours. >> I'm not sure what was done in the calculation, though, I assmue a lot of >> matrix-matrix, matrix-vector manipuation were performed. > > I wondered about this too. The article really doesn't touch on the > specifics of the type of calculations performed. But as I understand > most parallel computing needs are centered around matrix calculations > (as you said). > > I also assume that the article talks about speeding up Excel's worksheet > functions (aka cell functions), especially the matrix functions, instead > of some isolated computation routines such as solver. > >> I assume Excel might use BLAS, LAPACK and ScaLAPACK for such kind of things. >> >> I like GotoBLAS, the fastest BLAS and partial LAPACK implemntation, but >> it is not a free software. ATLAS is a free software though, it tunes itself >> while building. > > So, I know little about parallel computing or clustered computing, and I > do know that this is your area of expertise. With that in mind, I'd > like to ask you this question. How can an application make use of > clustered computers to perform parallel computations? Is there any > library that takes care of the complexity of sending/receiving > computation requests to multiple machines? > > It's sort of my pet peeves to make Calc more appealing to the scientific > community, but unfortunately I don't have enough expertise to follow > through with that. So, I would much like someone to educate me in this > area. > > Thanks, > > Kohei > > -- Erich Neuwirth, University of Vienna Faculty of Computer Science Computer Supported Didactics Working Group Visit our SunSITE at http://sunsite.univie.ac.at Phone: +43-1-4277-39464 Fax: +43-1-4277-39459 --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
