you're absurdly wrong regarding wii u and ps4.

OK, so I made a mistake on the Cell processor. I was thinking of double precision floating point performance for some reason. Even then, with only 6 SPUs available to devs it comes to 150 GFlops (6 x 25 GFlops). The theoretical 230 GFlops numbers is for the full 8 SPE Cell processor (I'm guessing the PPU contributes 30 GFlops so 180 GFlops total?).

As for the RSX, I'm not sure I believe that figure, partly because it comes from NVidia and is likely just marketing tripe, partly because the 7800GTX never topped 200GFlops (so 400 is absurd considering the similarities) and partly because in the same wikipedia link above is a link to the RSX article that states it's 176 GFlops (http://en.wikipedia.org/wiki/RSX_%27Reality_Synthesizer%27).

MK1 maybe had been very famous in Brasil but US and the rest of the world started talking about MK like today, after MK2.

MK3 was the first game to introduce a combo system and even then it was trounced by KI when it came to that. Johnny Cage was never worldwide icon, you might like him my friend but he wasn't that important. Actually he was omitted in couple of games in the series. Liu Kang, Scorpion and Sub-Zero on the other hand yes.

Why were Ed Boon and Tobias cult names? First time hearing that.

You should really read my post from site 7. Flop measurement is made always by the producer of that particular processor. It also has no value to compare flops from 2005/6/7 to todays flops because the theoretical maximums dont tell you anything and there were several architectural changes inbetween that makes the Chip handle game code way way way  better. The HD 5870 had 2.7 TFLOPS years ago does that mean it can actually do more than newer GPUs with less flops ? A HD 7870 GHZ runs circles around it despite having 200 less gflops. And they are just 2 gens apart.

Flops are not a metric for measuring graphics. 

Nvidia says its 400 gflop marketing bullshit or not what does it change if its just 200 gflop or 50 gflop ? How does a flop translate to graphic ? 

Flop comparisons make only sense in a gen of GPUs when flops are coupled with all other units and scale lineary. 

So if I say the PS4 is 3x PS3 it means nothing at all because its not comparable at all. PS4 might aswell be 20xPS3 depending on the workload.

How do you compare something based on theoretical flops that can massively vary depending on the code and all in the context of a nebulous quality like graphics. 

I based my math based on what card the PS4 is close to in performance (Actuall real world performance), and then compared that to the specs of the card in the PS4.  It is like an HD 6450 Vs. a 7870.  That is as big of a difference as I said.

I adressed the PS4 with a response to someone else, and LOL you Wii U people just won't get your heads out of the sand will you?  The Wii U has NO GAME that looks better than Killzone 2 or even BF3.  At best it will look slightly better but it will b e hardly noticable.  

PS3's theoretical max is 230.4 GFLOPS

http://en.wikipedia.org/wiki/PlayStation_3_hardware

I've had a loooong talk with CGI-Quality. PS4 is what 1.84 Terraflops? That's not quite 9x the power, but remember PS3's bottleneck was the RAM, which PS4 has plenty of.

They need to teach each other about Xbox One, before saying about it in public. There are too many blunders.

How much power is a gflop ?

I guess Mario Bros was 500kflop so 1mflop is 2 mario brothers or short Marios. 2000 Marios are 1gflop. 1 TFlop is 2000000 Marios or 2000 Giga Marios so PS4 is 3660 GigaMarios or 3.66 Marioflops. 

Maybe, I might be underspecing Wii U. How many GFLOPS can it push? From what I read Wii U is between 350-530GFLOPS. So at the minimum, PS4 with 1.84 GFLOPS is 3.5x as powerful as the Wii U. At a maximum it's 5x as powerful.

Also this would put Wii U at 1.5 to 2x as powerful as the PS3, so WiiU is still closer to PS3 then PS4.