How can I find the condition number of a large sparse matrix ?

I have a sparse matrix data which has 1)row index 2)column index and 3)non-zero matrix entries stored in 3 columns in a file. I am using 'spconvert' function to read and store the matrix in sparse format. The matrix size is 513263 x 513263 with density = 9.886x10^-6. But whenever I use 'condest' function to compute the condition number MATLAB stops, sometimes giving an error message that it is 'OUT OF MEMORY'. I used 'whos' to find out how much bytes it is storing. It turned out that it is storing 44943696 bytes approx 45 Mb. But whenever it computes through 'condest' function it stops after sometime. Why this is happening ? I've noticed one thing. When I ran this code on a Windows machine it was taking around 14 GB memory. Is there anything wrong with the matrix format ?

[Matt J]

Show us the output of the memory command when it happens,

   >>memory

[krishna chandran]

Its a GNU system not Windows.

>> >> >> >> >> >> >> >> >> >> {^HError using memory Function MEMORY is not available on this platform. }^H >> >> >> >>

[krishna chandran]

when using (A:)=(A:); >> >> >> >> >> >> >> {^HMatrix is too large to convert to linear index. }^H

[krishna chandran]

I've noticed one thing. When I ran this code on a Windows machine it was taking around 14 GB memory. Is there anything wrong with the matrix format ?