8-byte representation of a real number consisting of a sign bit, an 11-bit binary exponent, and a 52-bit mantissa, plus the implied high-order 1 bit
Note 1 to entry: Normalized double precision floating point numbers have a range of ±[1 + (1 – 2−52)] ∙ 21 023 ≈ ± 1,797 693 134 862 315 7 ∙ e308. The smallest value is 2,225 073 858 507 201 4 ∙ e−308.
Note 2 to entry: The less significant bytes are stored in lower memory addresses; the more significant bytes are stored in higher memory addresses.
Note 3 to entry: See IEEE 754-2008 for binary floating-point arithmetic.