> > So I would like to prove this observation by a good test. But > > comparing a retarded 64bit system with a full 64bit system isn't the > > solution. (Who would for example shut down half of the heating of his > > toaster anyway?) heres a test i did. compile glibc 2.3.6 (nptlonly) on 2 setups. athlon64/turion64, both single-core 2.0ghz connected via distcc over tcp over wifi. make -j4 time: 22 minutes macbook pro. 1.83ghz (32bit) dual-core. make -j3 time: 23 minutes seems like theres no real difference. except of course that if you run 64bit, you need to have your entire depchain down to glibc installed twice if you want to watch youtube or use PD :)