Arithmetic multiprecision
The multiprecision arithmetic indicates the whole of the techniques implemented to handle in a Computer program numbers (whole, rational, or floating mainly) of arbitrary size - in opposition to the numbers “machine” to which the operations provided by the processor apply. The size of which it is question is the number of figures used to represent the number: thus, into arithmetic multiprecision, it is hardly limited but by the memory available, while the arithmetic operations of the usual processors relate to entireties and the floating ones of to some computer words.
Many algorithms were developed to effectively carry out the usual operations on numbers comprising a very great number of figures. The algorithms of fast multiplication of great entireties are in the middle of this field. Indeed, of many more complex operations, to start with division, use the multiplication of entireties like building block, and the effectiveness of the algorithms used rests in an essential way on that of the subjacent multiplication.
On the technical plan, various libraries provides structures of data and operations effective for multiprecision calculation. Most widespread is probably GNU Multiprecision Library.
References
| Random links: | Sycorax | Esquelbecq | Fixed point | Colonel de Groulard | County of Wetzel | Mon_cerveau_accroche_upside-down_(Bonzo_va_à_Bitburg) |