I honestly don't know of any meaningful way to compare the performance of the human mind to that of a computer's. Each excels at things the other finds virtually impossible. The very best face-recognition systems will typically fail to recognize you somewhere 10-20% of the time, unless the conditions are basically staged - by contrast, when was the last time you really literally didn't recognize someone familiar to you? OTOH, a computer can calculate
pi to an absurd number of digits while you're still looking for pencil and paper.
Computers can handily beat all but the top one or two players in the world at chess - chess lends itself well to mathematical analysis, which is the computer's strong suit. The Japanese game of "Go" does not, however - even a novice Go player can generally easily beat the very best Go-playing computer.
We do some things well that computers don't, like pattern recognition, and computers do some things well that we don't, like mathematical analysis. That's about the only accurate comparison I think anyone can really make ;)
We're in the process of creating a silicon based life form.
I know that sounds tinfoilish but think about it. Think about how computers will progress if we survive another 500 years.
In his book, Age of Spiritual Machines,
Ray Kurzweil talks about how close we are to duplicating 'neural nets', essentially latice-like intelligence structures which mimic how the human mind works, and what we need, processor wise in order to duplicate it.
We'll be able to 'photocopy' a human neural net by the end of this century. Ostensibly, my two-month-old will never 'die' in the traditional sense. Her body will most likely die in about a century, however, her mind will probably have determined another place to call home.
This isn't sci-fi, this is odds-on likely. Of course, we still don't have flying cars, so who knows, but Kurzweil has done some amazing things.
There are some very sophisticated projects out there you can find on his site
http://www.kurzweilai.net.