I would love to see those benchmarks. Not trying to say you are wrong, I did some quick and non-thorough googling and saw a different story (From CPU boss, It looks like the score was related to a performance benchmark, but maybe they are simply going by the clock speeds, cores, etc...)
Cpuboss is a horribly innacurate tool as it's often based on benchmarks that aren't necessarily even consistent between components (check out an R9 295x2 vs. a GTX 780, the 295x2 is basically twice as powerful yet the 780 is listed as being superior due to the fact that benchmarks are missing). So you have to be kind of careful with CPUboss/GPUboss/etc..., and even synthetic benchmarks are something which need to be taken with a grain of salt as they're usually not very accurate representations of realistic workloads (Prime95 for example hammers a CPU far more than any realistic task ever will) and it's not unheard of for a benchmark to favor one competitor over the other (generally in favor of Intel more so than AMD) due to some underhanded tactics/and/or optimizing for Intel hardware.
Anyway, so AMD's FX series has similar IPC to Intel's C2Q's, but sine it clocks significantly higher you end up with single core performance in-line with first generation i7 (Nehalem), and multi core performance similar to that of Ivy Bridge. With the exception of rendering and virtualization tasks, almost everything else prefers stronger single core performance above all else and as such even a Haswell i5 (and in many cases an i3) outperforms AMD's FX series. For things like virtualization/rendering -- things that truly benefit from and can leverage the cores the FX series catches up quite a bit (and sits somewhere between an i5 and an i7 -- closer to an i7 though).
For gaming benchmarks, there are just too many to post, so if you do care then just look up CPU benchmarks for any game that comes to mind (I personally like to use Techspot as they usually will have a CPU performance comparing Intel's i7 at various clock speeds and AMD's FX8/9 at various clock speeds (so a 1230v3 would be virtually the same as a 4790k @ 3.5ghz for example since the Xeon E3's generally aren't benchmarked in gaming).
Now, you also have to keep in mind that a proper FX9590 setup (with a VERY high end board and a high end cooler) will end up costing about the same as an i7-based setup (so more than a 1230v3 setup) while consuming quite a bit more power and while yielding worse performance. And that's really the big problem with the FX series, in general it's similarly priced with Intel's competing offerings while being outperformed. Now, there are price points where the FX makes sense. An FX6 setup is similarly priced to an i3 setup, so for gaming it's kind of a toss up as to which to go for that's entirely dependent upon which games you play. An FX8 setup is similarly priced to a locked i5 setup, so for gaming the locked i5 is better (unless you plan on streaming), but for rendering/virtualization, the FX8 is the better choice (until you up your budget a bit more and can afford a 1230v3-based system).
Like I said before, there are gaps with benchmarks, and not all of them cover all the relevant CPUs, so some assumptions need to be made based on the information that's available.
http://www.anandtech.com/show/8427/amd-fx-8370e-cpu-review-vishera-95w/2
Ultimately, a lot of what I've seen/heard comes down to anecdotal evidence where people, whom I would consider to be a reliable source (mods/people who have proven themselves on forums/etc...), who have switched from an FX cpu to an i7/Xeon have said how their workflow was affected and how even just working with/navigating in Windows itself feels snappier.
3901, 14936 does seem to beat
2833, 14499
especially in the per-core department which tends to matter for NAS. Overall I think the
"outperform ... across the board" might be a bit of a stretch, click that for what I suspect is a fairer comparison.
It depends what you consider outperforming. In my mind, if it's faster (more than margin of error), then it's outperforming. In some cases the difference can be massive (like in many games), while in others (rendering/virtualization/some other games) the difference is far less significant. So would I recommend someone upgrade from an FX9590 to an i7 -- mostly no, but there are some cases where I would, but I also generally wouldn't recommend someone buy an FX CPU (never an FX9590 as it's just an overclocked/highly binned 8320 that costs too much considering it's performance and power consumption).
