4 ms·
ARMv6 doesn't support hard float, right? Why would you want to do calculations on an Pi?
by misframer 13y ago
ARMv6 doesn't support hard float, right? Why would you want to do calculations on an Pi?
- justin66 13y agoRaspberry Pi has an FPU.
- tokipin 13y agoMathematica can deal with arbitrary-precision numbers (or better said, Mathematica doesn't blindly use whatever underlying number format is present, it has precise controls for this sort of thing http://reference.wolfram.com/mathematica/tutorial/ArbitraryPrecisionNumbers.html http://reference.wolfram.com/mathematica/tutorial/ArbitraryP...)
- Narishma 13y agoThe ARMv6 CPU in the Pi has an FPU. Depending on your OS, you can use it in two different ways: * softfp where the float function arguments are passed in using integer registers or the stack * hardfp where they are passed in using the FPU registers In the above cases, the float operations are done in hardware in the FPU. There is a 3rd ABI called softfloat where everything is done in software, and thus is very slow.