Most current languages don't 'handle' floating point numbers themselves, they simply use the CPU instructions available to work with them. The only way around

Most current languages don't 'handle' floating point numbers themselves, they simply use the CPU instructions available to work with them. The only way around that is to use a library that supports higher-precision floating point through emulation. Then agian, that still doesn't change the theoretical background in any way.