What disadvantages are there in using decimal arithmetic?
Decimal numbers are traditionally held in a binary coded decimal form which uses about 20% more storage than a purely binary representation. More compact representations can be used (such as the Densely Packed Decimal and Chen-Ho encodings) but these do complicate conversions slightly (in hardware, compressing or uncompressing costs about two gate delays). (Pure binary integers can also be used, as in the Java BigDecimal class, but this makes rounding, scaling, conversions, and many other decimal operations very expensive.) Even with these more efficient representations, decimal arithmetic requires a few extra bits (an extra digit) to achieve the same accuracy as binary arithmetic. Calculations in decimal can therefore require about 15% more circuitry than pure binary calculations, and will typically be a little slower. However, if conversions would be needed to use a binary representation because the data are in a decimal base then it can be considerably more efficient to do the calcu
