I was wondering if somebody could give me a technical breakdown of why a 2GHz Octa Core is faster than a 4GHz Quad Core in multithreaded workloads, assuming both chips has the same architecture and are of same generation with all the same specs and IPC, when they have the same amount of total GHz available? In theory a 4GHz Quad Core should perform equally against a 2GHz Octa Core if the program is optimized for utilizing both 4GHz and 8 cores? Or am I missing something here?
We would have to look deeper into this inquiry, from your post I see a couple of assumptions that can actually affect the results with these platforms. If you have the models, I can help by comparing them and find a possible reason of why there could be a performance difference.
I didn't have any two Intel/AMD chips in mind per se, I was phrasing my question more as a hypothetical one and was wondering about the differences of the two hypothetical chips mentioned above and a technical breakdown of the differences in terms of multithreaded workloads. I could be asking the wrong channel in the Intel forum, but this was the closest channel I found.
/thread/112690 KernelXanders, from my end the best support I can provide is to help comparing features between to Intel products, and possibly get a reason of why there could be a performance difference. I believe that your question is related to benchmarking which in that the case the best places to get more answers are third-party sites. Places like; http://www.passmark.com/forum/ Forums - PassMark Support Forums and also http://www.anandtech.com/bench/CPU/39 CPU Benchmarks - Compare Products on AnandTech where they take similar products test them and share their results.