LongMath is an extended precision math package for the HP48SX. It should work on the GX as well, but has not been tested. Let me know if it does or does not. It handles negative numbers and decimal points with precision limited only by free memory. Addition, subtraction, division, multiplication, modulus, negate, conversion to/from binary, fibbanachi and factorial routines are included. The most time consuming routine has been written in assembly for speed, and other often used programs in SysRPL. It is still somewhat sluggish, but is still useable. Division is the slowest.
The current version is V1.1 written in April of 1994. You can get the LongMath.zip archive here via HTTP with "Load to Disk" set on your Web browser.