128 bit Data Type

[abc876]

I want to perform some calculation which require extremely large floating point values i.e. > 1.7e308. I want some class for handling such large calculations. 128Bit or more. Any tips would be welcomed. Thanks! Muhammad Shoaib Khan http://geocities.com/lansolution

[G Haranadh]

May be you can get some help with this link. :-D http://www.codeproject.com/cpp/VisualCalc.asp

[toxcct]

actually, my VisualCalc 2.13 uses simple doubles for now... if he had wanted 128 bits integers, he could have used __int64, but for floating point numbers...:wtf: i really don't know.

---

TOXCCT >>> GEII power
[toxcct][VisualCalc]

[Bob Stanneveld]

Hello, You could always create a class that holds a memory block of 128 bits. You have to decide which floating point representation you use and define all the basic operations yourself. You have to go down to the bit level yourself to do all the excact calculations. One class declaration could look like:

class CFloat128
{
public:
CFloat128(unsigned __int64 uLowOrder,
unsigned __int64 uHighOrder = 0,
bool bNegative = false);
~CFloat128();

// your operations
private:
void* m_pData;
};

// overloaded operators
CFloat128 operator + (const CFloat128& lFloat, const CFloat128 rFloat);
// etc..

Hope this helps :) Behind every great black man...             ... is the police. - Conspiracy brother Blog[^]

[markkuk]

GMP[^] supports arbitrary precision floating point math.