floating point

Source board

This definition comes from the board below. Open it to see the term in context.

Numbers for RealHow Computers Work

More terms from this board

• epsilon compare|a-b| < tiny instead of ==
• IEEE 754the standard (float32 = 1/8/23, float64 = 1/11/52)
• Infoverflowed-to-infinity
• mantissathe significant digits
• modular arithmeticclock/wheel wraparound
• NaN'not a number', ≠ itself
• overflowa result too big for the width, wraps around (mod 2^n)
• rounding errora value not exactly representable in binary (0.1)
• signedallows negatives
• the 2038 problem32-bit Unix time overflow
• two's complementnegate by flip-bits-then-+1; one adder does both; range -2^(n-1)..2^(n-1)-1
• unsignednon-negative, 0..2^n-1