Timing processes across different architectures

Hi All:

Given an arbitrary computational process run on an arbitrary set of
hardware, and the time needed to complete the process, is it possible
to extrapolate how long the process would take on another set of
hardware?

Example 1: A process takes 10 seconds to complete on a P4 3.0 GHz with
1 GB of RAM. How long will this process take to complete on a Centrino
1.8 GHz with 512 MB of RAM?

Example 2: A process takes 24 seconds to complete on a Xeon 2.4 GHz
with hyperthreading enabled, and 2 GB of RAM. How long will this
process take to complete on a P3 900 with 256 MB of RAM?

I know there details about the particular process (CPU intensive, disk
I/O intensive, etc.) to take into consideration, but I'm interested in
a very crude approximation. Is there a simple equation, or a graph
curve available that I could look at and estimate performance
improvements/degradations across different sets of hardware?

Thanks for any help.