If we were to reflect back upon our results during our initial review of the “Prescott” core, we would see that initially, the SANDRA benchmarks were not too kind to the “Prescott” core, mainly due to the incredible pipeline stage increase. Now that we have some additional results, we can get a better idea as to what is going to happen in the future.

     If we take a look at the Arithmetic results, even though we see that the “Prescott” is once again lagging behind an equally clocked “Northwood”, the general slope of the ALU results appear to outpace the growth of the “Northwood”. In the long run, this means that even though the pipeline was increased, the other additions and changes to the architecture that Intel has made will allow for “Prescott” to ramp performance faster that “Northwood”. For our FPU results, we seem to see the opposite with the “Northwood” core posting a better performance increase per MHz jump.

     As we progress on to the Multimedia results, we once again see that the “Northwood” is easily able to best the “Prescott” due to the pipeline increase. We also again notice that the ALU performance slope for “Prescott” is higher than that of the “Northwood” while the FPU results show a slightly slower growth. I was hoping for some far better Multimedia marks coming out of the “Prescott” core, a thought that I noted in during our initial review. With all of the added cache, and architecture changes geared towards performance enhancement, I hoped that “Prescott” would be better able to keep up with “Northwood”, but once again we see the perils of a deepened pipeline.

     The Memory benchmark results show some of the improvements that come with the new “Prescott” core. Here, we can see that “Prescott” is able to outshine “Northwood, although only by a small margin all things considered. “Prescott” shows a larger performance growth rate here than does the “Northwood” which is expected due to the increased on-die cache, among other memory performance enhancements made by Intel.