My apologies for writing in English, but my French is not adequate.
Mathcad represents numbers internally using the IEEE standard 64 bit floating point. This is a format that is standard in the industry, and which is implemented with high speed hardware in all processors since the 486 line. This provides for high speed implementation of complex algorithms.
Computers are not limited in their number represenations, and one can easily write routines to do arithmetic to any desired precision, limited only by available memory and patience. I do not know what format is used by the MS calculator. They could be using some unique format, possibly decimal rather than binary based (most pocket calculators use a decimal representation), or they could be using the IEEE 128 bit format. Either would be much slower than the IEEE 64 bit format, even if implemented in hardware (some processors could have the 128 bit format built in).
Mathcad does include a symbolic processor, very ditinct from the numeric processor. In versions lower than 14 this processor is based on Maple. In the symbolic processor (activated by using the symbolic evaluation function, &rarr) large numbers can be used. The floating point is decimal (rather than binary) based, resulting in rather different roundings, and results that generally better match hand calculations. The limitations are that it is very slow, and that while general mathematical operations and functions are provided, more complicated numeric algorithms are not. Depending on what calculations you need to do the symbolic processor may be adequate.
__________________
� � � � Tom Gutman
