[Numpy-discussion] Fast threading solution thoughts

David Cournapeau david@ar.media.kyoto-u.ac...
Thu Feb 12 06:48:58 CST 2009


Sturla Molden wrote:
> On 2/12/2009 12:20 PM, David Cournapeau wrote:
>
>   
>> It does if you have access to the parallel toolbox I mentioned earlier
>> in this thread (again, no experience with it, but I think it is
>> specially popular on clusters; in that case, though, it is not limited
>> to thread-based implementation).
>>     
>
> As has been mentioned, Matlab is a safer language for parallel computing 
>   as arrays are immutable. There is almost no need for synchronization 
> of any sort, except barriers.
>
> Maybe it is time to implement an immutable ndarray subclass?
>   

Is it possible at all ? I mean immutable array as a subclass of a
mutable array ? I too would like some immutable arrays  but this sounds
like an enormous task (as much as I like numpy, I find numpy sometimes
harder to use than matlab for some problems because of the immutability
- may be my own limitation, though),

> As for Matlab, I have noticed that they are experimenting with CUDA now, 
> to use nvidia's processors for hardware-acceleration.

Labview too; there was a small boost from NI at last ICASSP, and they
were demoing some GPU based operations (mostly blas/lapack/fft from what
I understood, the easy stuff).

cheers,

David


More information about the Numpy-discussion mailing list