This is to let you know that the Kappa Library may now be used to process star schemas for parallel and OpenMP CPU processing and/or into a GPU for CUDA processing. OLTP schemas may also be used. This allows processing and working with hypercubes using massively parallel processing (supercomputing) with commonly available datacenter servers. Boundaries of parallel execution are dictated by algorithm data flow which results in more parallel execution performance than other solutions. Kappa implements producer/consumer data flow scheduling and index component (tensor) notation expansion for fully scaled dynamic parallel execution based on current dataset dimension categories.
Kappa has language bindings and examples for .Net (C# and Visual Basic) as well as C++, Perl, and Python.
Take a look at http://psilambda.com and make sure to look at the Quick Start Guides. Feel free to ask your OLAP vendors for their plans for incorporating the Kappa Library or similar capability into their products.
Advertisements, add-ons and other commercial information
1 post • Page 1 of 1