OK, so where can I compare to other Athlon x2 users?
Search for forums. User results are sprinkled here and there. This WW benchmark was the first formal CPU benchmark that anyone attempted (to my knowledge). The only people that really have a good grasp on how many different chips will perform are the people that have been very active on these forums for a very long time.
For the most part you just look up what the architecture is, check out some single-threaded benchmarks and compare it to what you have seen other users report and then you give a verdict. Look up the Athlon x2's single threaded performance compared to an Ivy Bridge with the same clock rate and you will cry.
It can't be just any single threaded application. It has to be an application with similar logic to dolphin. The closest thing out there is a c++ compiler, but unfortunately we don't get many compiler benchmarks these days.
Cinebench seems to do pretty well. But the IPC differences in dolphin are still more extreme than cinebench (is extreme really the best word I can think of here? Jesus I'm tired).
That is true, but on a very basic level single threaded benchmarks roughly relate to Dolphin performance (although Dolphin does work differently than the benchmarks). It is more accurate than looking at multi-threaded benchmarks like geekbench.
Ok instead of reiterating my point with a well thought out paragraph why don't I propose a not so hypothetical question.
There are single threaded benchmarks where an FX 8150 beats a phenom II 975 and there are single threaded benchmarks where the phenom II 975 beats the FX 8150. How do you determine based on those benchmarks which cpu is better for dolphin?
(10-15-2012, 09:12 AM)NaturalViolence Wrote: [ -> ]Ok instead of reiterating my point with a well thought out paragraph why don't I propose a not so hypothetical question.
There are single threaded benchmarks where an FX 8150 beats a phenom II 975 and there are single threaded benchmarks where the phenom II 975 beats the FX 8150. How do you determine based on those benchmarks which cpu is better for dolphin?
You don't, but you do recognize that the performance difference between them isn't as large as the difference between an Ivy Bridge and either. That's why it is a very rough estimate. It doesn't distinguish between speeds for processors that are very close in single threaded performance really.
(10-15-2012, 10:13 AM)NaturalViolence Wrote: [ -> ]Hint: There is a way.
That requires much more effort than I am willing to exert
(10-15-2012, 10:29 AM)Axxer Wrote: [ -> ] (10-15-2012, 10:13 AM)NaturalViolence Wrote: [ -> ]Hint: There is a way.
That requires much more effort than I am willing to exert
Ok let me ask you this question then. How would you define or categorize dolphins code profile/logic? What distinguishes it from other types of applications? Surely we can find common ground with other applications out there.