Intel(-type) Processors IIRC use 80 bit FP numbers for internal calculations and do a conversion back to 64 bit when saving the results to registers.
If by registers, you mean the regular (non-FPU registers) this is correct. However the FPU register stack maintains values in 80 bit form exclusively, and these values CAN be copied to memory intact as 80 bit values. Parts of both statements may be right or wrong. The FPU internal computing registers are 80-bits but it also has a few 16-bit registers for various functions. One of them is the Control Word register in which 2 of the 16 bits are used for Precision Control and define how the results will be rounded after an instruction is performed. The choices are 32-bit, 64-bit or 80-bit. Depending on your OS, those bits may be set initially to 64-bit mode when a program first starts (such as under Windows). If those bits are not reset to get 80-bit precision, all results get rounded to the 64-bit precision when "saving the results to registers" although all computations are performed with 80 bits. The FPU cannot save any data directly to the regular non-FPU registers (and they're still only 32-bit registers in most computers). But, yes the FPU can copy 80-bit values to memory intact as long as the Precision Control has been set accordingly. However, if you store intermediate results as 64-bit values, you immediately waste the higher precision of the 80-bit computations. Raymond _________________________________________________________________ Help STOP SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=features/junkmail