Hi Maho,

On Friday, 2009-11-20 12:32:09 +0900, Maho NAKATA wrote:

> My idea is just a not a special and very simple.
> 
> If we do vector-vector, vector-matrix, matrix-matrix multiplication etc, 
> 
> *just call BLAS and LAPACK*

We don't do much of these operations, only in some context of
spreadsheet scientific functions and matrix functions of course.

Most time is spent querying search expressions and accessing cell values
and calculation, these operations and all document accessing functions
that can be called from within the interpreter have to be made reentrant
and thread safe. The mechanism that currently builds the calculation
graph (it isn't really one) on the fly as it encounters the cells to be
calculated will have to be changed to pre-calculate independent branches
within the graph that then can be fed to some threading engine, or later
MPI or some parallelizing framework.


> If you need parallel version, then call PBLAS, BLACS and ScaLAPACK. We may
> also need MPI for that.

> 4. We may need FORTRAN compiler as well :-)

I think you just lined out why we would not start with this approach
first. With general thread-safety we would gain more. My tuppence.

  Eike

-- 
 OOo/SO Calc core developer. Number formatter stricken i18n transpositionizer.
 SunSign   0x87F8D412 : 2F58 5236 DB02 F335 8304  7D6C 65C9 F9B5 87F8 D412
 OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS
 Please don't send personal mail to the e...@sun.com account, which I use for
 mailing lists only and don't read from outside Sun. Use er...@sun.com Thanks.

Attachment: pgpuQMlyTiptq.pgp
Description: PGP signature

Reply via email to