QlikView is using IEEE double-precision binary floating point. You cannot increase that precision.
You may be able to do better with particular calculations. I can't give you any advice without knowing the calculation and the numbers you're using, though.