If the benchmarks differ by as little as 10% then for all intent the CPUs are equal. The TOTAL PACKAGE weighs heavily on the final benchmark.
What good is a sizzling hot fast CPU if you tag it with slow graphics and a 4200 RPM hard disk?
Step back and benchmark the entire unit.
How can I compare two CPUs to see which one is faster? I know the answer to this will probably depend on the type of application/benchmark used, however I would guess that in most cases you can decide which one is generally faster.
The reason I'm posting this instead of checking benchmarks for myself is that I am confused by these two charts (the only ones that I found to actually rank several CPU models in the same searchable list):
The two lists seem to contradict each other quite often. For example, in the first list, the Intel Pentium Dual Core T3200 (rank: 91) seems to be well slower than the Intel Core2 Duo T5800 (rank: 56), while in the second list, T3200 even gets a higher passmark (1049) than T5800 does (1002).
Also, looking just at the first list, it seems the T8300 (rank: 28) is faster than the T8400 (rank: 35). It would make sense that the higher model number would also be faster. I'd be prepared to trust these charts that this isn't actually how Intel names their products, but I'd expect to see more or less similar results in different charts.
So which of these two (if any) should be trusted?