Support for 64-bit processors has been a project for Pd in the past few years. We are getting very close to complete 64-bit compatibility, and this means we can move on to more exciting aspects of 64-bit processing. For Pd, replacing the 32-bit floating point number format with a 64-bit floating point format means that the single number type in Pd could also represent 32-bit integers that are very common, like UNIX timestamps and much more. Currently, the 32-bit floating point number format can represent about 19-bits of an integer.
Here is a rough history of 64-bit support in Pd:
first usable 64-bit builds for Pd vanilla
all major known 64-bit bugs fixed in Pd vanilla
first usable 64-bit builds for Pd-extended on GNU/Linux
all major known 64-bit bugs fixed in Pd-extended on GNU/Linux
complete 64-bit support for Pd-extended on Mac OS X except Gem
support for using 64-bit double numbers as the base number, see pd-double