John J. McDonough wrote: > Olin Lathrop wrote: > > John J. McDonough wrote: > >> There are some quirky things about the PIC30, maybe even quirkier than > >> the PIC16. > > > > Like what? My general feeling is the opposite. > > Like ... 16 bit registers, except for the 40 bit ones (why 40?) Most chips that have a one-cycle hardware multiplier intended for DSP operations (primarily multiply-accumulate, a.k.a. dot product) have one or more registers that are intended to hold the output of the multiplier without rounding or truncation. These registers are usually 2*N + 8 bits wide, where N is the base word width of the processor. On a 16-bit processor like the dsPIC, that works out to 40 bits. Such a register can accumulate the results of up to 256 multiplications without any truncation or overflow whatsoever, allowing the inner loop to run at its maximum speed without having to worry about exception handling. Any fixups or exceptions happen just once, after the loop is complete. -- Dave Tweed -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist