128 bit Decimal Type

[OriginalGriff]

How do you think 8 bit processors handled 16 bit integers? It's not difficult: you just have a half carry in the middle!

0x1234 + 0x6743
is worked out in 8 bits as:
0x34 + 0x43 == 0x77 plus a zero carry
and 0x12 + 0x67 + 0 == 0x79

== 0x7977

It's not done quite that way with decimals (because floating point arithmetic is a lot more complex) but that's the general idea. Think about it: you are a base ten processor: but there is no limit on how big a pair of numbers you can do arithmetic with!

You looking for sympathy? You'll find it in the dictionary, between sympathomimetic and sympatric (Page 1788, if it helps)

[PIEBALDconsult]

You had me at

OriginalGriff wrote:

how big a pair

[PIEBALDconsult]

See also: http://msdn.microsoft.com/en-us/library/system.numerics.biginteger(v=vs.110).aspx[^]

[OriginalGriff]

Clubs? Spades? :InnocentWhistleSmiley:

You looking for sympathy? You'll find it in the dictionary, between sympathomimetic and sympatric (Page 1788, if it helps)

[PIEBALDconsult]

At clubs usually. As for the other, I hesitate to use such language.

[OriginalGriff]

Grief! That's a term I hadn't heard or thought about for decades. I assumed that died in the 70's!

You looking for sympathy? You'll find it in the dictionary, between sympathomimetic and sympatric (Page 1788, if it helps)

[PIEBALDconsult]

Yeah, it's not used so much anymore. It appears in "Streets of Fire" (circa 1984) and if I recall correctly, Bill Maher used it a while back.

[Lost User]

It's not a type directly supported with processor instructions dedicated to it. You can look at its source[^] to see how it works (does not include the source of some math operations and so on, but shows their function names).

[Ravi Bhavnani]

You may be confusing addressing space with data size. A 64-bit computer can address 264 bytes.  An instance of a 128 bit decimal type (located somewhere in the address space) occupies 16 bytes. /ravi

My new year resolution: 2048 x 1536 Home | Articles | My .NET bits | Freeware ravib(at)ravib(dot)com

[jschell]

computerpublic wrote:

Somehow the logic doesn't seem to come to me

Might help if you wrote your own class to do infinite digit integer handling. So you want a class that can take a 50 digit integer and add a different 50 digit integer to it. Each class instance represents a single number. So the add operation would look like the following. MyInt op1 = new MyInt("1000....001"); MyInt op2 = new MyInt("9000....019"); MyInt result = op1.add(op2);