Yeah, why wouldn't we let the guy screw us like he always did one more time and shut the :bananal: up about it :rolleyes:
Printable View
then this thread should be closed.
Could you update this chart and add two columns. Next to the Phenom 2600 results add a column with the Phenom I normalized to 3 GHz, and the second column a Phenom II at 3 GHz (I realize it will be empty until the NDA expires). This will give a good comparison of the performance increase in the new cpu's.
I think this chart is very interesting from two standpoints. You can see a consistent pattern. For applications in which the Phenom performs poorly with respect to the QX9650 quad or the core i7 920, the increase in performance seems to be more of a function of the large cache available. For other applications, the Phenom 2600 either wins or is very close, despite being at a lower speed. I think we will see a rapid convergence of scores when the larger cache Phenom II's become available. I am beginning to think that most of the original Phenom's problems have been that it has too small an L3 cache for four cores and poor cache latency.
Of course I could be wrong...it seems to happen a lot :rofl: