What is Subnormal number?

A floating-point number whose exponent field is all zeros, without an implied leading 1, representing values smaller than the smallest normal number. It fills the gap to zero gradually (gradual underflow), so x − y = 0 implies x = y.

Field: Computer Arithmetic

Related terms

Arithmetic shift bfloat16 Booth's algorithm Carry-lookahead adder Carry-save adder Catastrophic cancellation Exponent field Fixed point Floating-point absorption FP8

Back to the glossary