nrd said:
I recently sold my 2.0 G5 PowerMac (dual chip, not dual core) for a 2.0 MacBook Pro. I notice no significant difference between the two machines. I'm sure, however, that testing of real world use would likely show that the G5 is slightly faster because of two reasons: faster FSB (and thus memory) and a faster (or should I say less latent) hard drive.
I'm sure if you did integer/floating point tests, one chip would excel over the other, but for most purposes, the machines are very similar.
Unfortunately, the 2.0 GHz G5 has
slower memory than the MacBook Pro. The front side bus (the connection between the processor and the 'northbridge', the chip that connects processor, memory, video chip, and 'rest of computer',) is slower, but the memory is faster.
Your dual-processor 2.0 GHz Power Mac G5 has a 1 GHz front side bus, per processor, but only 400 MHz DDR memory. This means that the processors can talk to each other significantly faster than they can talk to main memory.
The Core Duo, on the other hand, has a 667 MHz front side bus (technically it is shared between the two cores, but the two cores are linked internally even faster than the G5,) with 667 MHz DDR2 memory. In all likelyhood, the Core Duo's memory is higher latency than the G5s (DDR2, in general, is higher latency; but if Apple skimped on the G5, they
could be the same,) but faster transfer speed. This applies when memory is installed in pairs. The G5 forces dual-channel memory, whereas the Core Duo supports either single-channel or dual-channel mode. The Core Duo would be slower with just one memory module installed. (Which is the default configuration. You have to build-to-order, or add after purchase, to get dual-channel.)
Because of the combination of (potentially) faster memory, and faster inter-processor communication, but slower processor-to-northbridge connection, it probably all balances out.
