AMD users often cite the PassMark benchmark as "proof" that AMD's Bulldozer trumps Intel's Sandy Bridge, particularly the 8120 versus the 2500k.
I've never taken passmark seriously, nor have I to my knowledge ever used it as a basis for any claims I have made on these forums.
Secondly, the performance of each core is tallied up as a whole, and the FX-8120 has eight of them. The 2500k has only four, and with no hyper threading. That is clearly relevant seeing as how the 2600k, with the same exact architecture as the 2500k, out performs the 8120 handedly, despite the fact that four of it's cores are virtual. And overclocking the 2500k would leave the 8120 in the dust even more so.
I gave a link that shows an 8120 up against a 2500K under overclocking conditions that covers many applications, including synthetics and actual gaming. Its clear to me based on those results that the 8120 priced at $170 can clearly stand up to the $220 2500K regardless of how much higher the results
should be. Dealing in hypotheticals is kind of pointless, of course the architecture is not as efficient as it should/could/ought to be, I don't dispute this, but lets stick to facts. Furthermore, the 2600K is a $300 dollar CPU, I darn well expect it to perform better than a $170 dollar one, regardless of who makes it or how its made. I try to keep my claims solely based on Price vs Performance comparisons only, and include overclocking potential when applicable.
To each their own of course applies. But just in case you missed it, here it is again:
http://www.overclock.net/t/1210060/fx8120-vs-2500k-benchmark-results