Parallella Community

Hello,

I am running "Fast Independent component analysis" on a windows machine which is i3 and a 64 bit OS. Right now FastICA takes about 630 milli seconds.I want to know what will be the speed on the parallela. If I have to print it down to 500 microseconds, what would you recommend (which board and how to parallelize it etc)?

How do I do matrix multiplications and other mathematical operations on the parallella? Could you point me to the exact example code? (what library to use and example?)

How about this? - https://github.com/parallella/parallell ... sembly-opt

You might also want to look at this - https://github.com/parallella/pal