[log in to unmask] wrote:
>
>
> Yes, your use of the word "significantly" does indeed answer my question.
>
> As I said previously, I am using an Alpha box (VMS) and finding that the
> difference between single and double precision is no longer as much a time
> problem as used to be. Our programs are now running in minutes rather than
> hours/days several years ago. Changing our programs has enabled us to consider
> larger (less numerically stable in single precision) problems, and our real
> world has become larger over the last few years (interconnection of more states
> in the Australian electrical environment).
>
> Cray's single is the size of VMS double, so I should be comparing quad on my
> machine with your double (a recent thread on c.l.f)
>
Correct. Both Cray and Alpha (and most, if not all, today's CPU) internal floating-point units are working on 64bit floats. So calculations on 64bit floats (REAL on Cray, DOUBLE PRECISION on Alpha) are generally as fast as calculations on 32bit ones (which are actually performed on 64 bits, and disregarding memory transfer rates) and much faster than calculations on 128bit ones (software emulated).
Pierre
--
+-----------------------------------+----------------------------+
| Pierre Hugonnet | mail....CGG |
| | 1, rue Leon Migaux |
| R&D Data Processing | 91341 MASSY cedex |
| | FRANCE |
| COMPAGNIE GENERALE DE GEOPHYSIQUE | phone...(33) 164 47 45 59 |
| Paris Processing Centre | fax.....(33) 164 47 32 49 |
| http://www.cgg.com | [log in to unmask] |
+-----------------------------------+----------------------------+
My opinions are not necessarily those of CGG
--------------------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|