Multiplying floating-point numbers by floating-point numbers in C++, does the final precision change?
The accuracy varies slightly. Generally speaking, the binary bits used when floating-point numbers are stored in a computer are fixed. What is the original precision of mantissa and what is the multiplication? The accuracy remains the same. In extreme cases, the last bit of the product mantissa may not be a significant bit (every bit of the factor mantissa is a significant bit)!