Main Page | Report Page

 

  Computers Forum Index » Computer Languages (IDL-PVWAVE) » SVD on CUDA...

Author Message
Mort Canty...
Posted: Sat Oct 23, 2010 5:35 pm
 
After two days fighting with IDL arrays vs. matrices, column major vs.
row major, transposition, leading dimensions and my own stupidity, I
finally got a DLM to encapsulate the CULA Tools single precision SVD
function. On my antiquated Nvidia GPU I get a sevenfold speedup relative
to LA_SVD for a 1000x2000 matrix. I'm asking Santa for a Fermi :-)

If you're interested, see

http://fwenvi-idl.blogspot.com/2010/10/singular-value-decomposition-on-gpu.html

Cheers,

Mort
 
 
Page 1 of 1    
All times are GMT
The time now is Thu Aug 21, 2014 9:57 pm