floating point - Does Java's BigDecimal leverage the hardware architecture like long double in C++? -
as understand it, long double in c++ leverages hardware architecture (at least architectures). bigdecimal in java small enough inputs?
does bigdecimal in java small enough inputs?
no, , not. floating-point numbers lossy, while every bigdecimal
has associated precision, , can represent number below precision. there no "small enough input" can usefully rendered floating point number, because if happen have bigdecimal
value can represented in floating-point notation, you'd hard-pressed sort of operations on value , maintain specified precision.
put way, purpose of bigdecimal
offer precision while sacrificing speed. runs contrary floating-point, sacrifices precision in favor of speed.
it sounds you're asking if java offers way work long double
sized floating-point numbers, , there not. can conclude fact it's not provided jdk authors have never felt necessary add language.
Comments
Post a Comment