double-precision number

double-precision number

[¦dəb·əl prə¦sizh·ən ′nəm·bər]
(computer science)
References in periodicals archive ?
In contrast, IEEE-type quad-precision numbers might have 113 bits of mantissa; a double-precision number has 53 bits; and a double extended number has 64 bits of precision.
In this representation, a double-precision number is represented as the sum of two singles.
More specifically, we assume that a double-precision number has 53 mantissa bits in memory and 64 in the registers.
Note that we do not use the fact that the inputs are double-precision numbers.
One way to get the high and low parts of the product of two double-precision numbers on a machine that supports quad-precision variables.
One way to get the high and low parts of the product of two double-precision numbers on a machine that does a fused multiply-add but does not support a quad datatype.
One way to get all the digits of the product of two double-precision numbers stored in single-single format.
We also assume that the machine has an instruction that returns the quad-precision result of arithmetic operations on two double-precision numbers.
If we are using normalized, IEEE floating-point, double-precision numbers which have an implicit leading 1, mantissas lie in the interval [1,2).
In this case we assume that our quad-precision numbers are stored as two double-precision numbers each having 53 mantissa bits in both memory and the registers.
Method Component Total Bytes U km double-precision numbers SVD V kn double-precision numbers 8k(m + n + 1) [Sigma] k double-precision numbers X km numbers from {-1, 0, 1} SDD Y kn numbers from {-1, 0, 1} 4k + 1/4k(m +n) D k single-precision numbers
Numeric variables can be convened to string variables by using the MKS$, MKI$, or MKD$ functions for single-precision numbers, integers, and double-precision numbers, respectively.