If the PS4 could do 10 years (I doubt it) yes defiantly, why not... But I don't think this generation will have the same leg as previous generations. I'd say that maybe it can beet PS1 total sales by the end of the 8 generation.

Sony is the primary reason console gaming is as popular as it is today. Regaining dominance in typical form will widen the market back to the way it was previously. Perhaps they will relinquish  twenty or so million consoles to their competition more than gens in the past, but its settled that Sony is back in typical dominant form. 

I don't get why people think this will be a short generation. I think it will be at least as long as Gen 7. The tech just doesn't evolve as fast as in the past. CPUs a re pretty much stagnating and considerably faster GPU units are so damn expensive that they aren't interesting at all for a console. On top of that, you don't see as much difference as back in the day from more hardware power. The resolution might go up a little, the fps might go up a little, but that's pretty much it. If there was a new console in 2 or 3 years, we would be seeing Cross-Gen titles like there's no tomorrow, because every game could be easily ported to the last gen. With that happening, many many people will just stick to their PS4/X1, because why should they throw out another 400+ $ when the old console is doing just fine and is still getting all major games? It would be a pretty high risk for any manufacturer to release a console too early.

New consoles will arrive when they are needed and the most important thing for that are significantly faster APU units. They need to be at least 5 times faster, better yet would be factor between 8 and 10. An APU like that isn't even on the horizon yet and you sure as hell won't be seeing a manufacturer packing a dedicated GPU into a console. The price would be insane, same as the power usage. Not interesting for the mass market.

Moore's law is pretty much gone and that's why I think console generations will only get longer and longer from here on out. I don't expect a PS5 before 2020.

Your prediction is impressive, only time can tell but one thing. When you write millions, you don't need the "S" at the end. It should be 140 Million or 140 Million Units. Million only needs to be plural if you're saying something like, "Millions of Ps4's are selling every year". Just wanted to help. I have grammar mistakes also but would rather them be corrected.

I agree with you, but a few thing could change this, per example if VR take off, microsoft want be parts of this and get a new console with more power to get good VR 1080/60fps and sony would just follow, or simply microsoft Xboxone start to sell bad and like nintendo get their console out earlier, i cannot see sony not follow microsoft and lets them have a headstart !

Just because you are right so far doesn't mean you'll be right in the end

Personally, my guess is the PS4 will sell 120 million by the time it's discontinued. I doubt it can get as close to the PS2 as it would be by selling 140 million.

CPUs are stagnating since there isn't any pressure on Intel right now from AMD (may change with Zen) or IBM (at servers) to develop any faster. Intel is absolutely dominating the market, with the exeptions of the low power mobile and embedded markets where ARM shines.

As for GPU, the main problem had been the problems with the 20nm process being barely any better than the old 28nm process. Which is why they where stuck there for 4 years now. 14/16nm will give a big boost to mid tier GPUs (those in range both financially and in TDP for consoles), and architectonal changes will further boost them. And by then mid tier will be former highest end performance, like former Hawaii or GTX 780 like range of performance.

Also, with the bottlenecks in drawcalls and bandwith being dealt with, GPUs can continue to grow at a fast pace.

About the APUs in PCs being much smaller, that's true. But the reason is not the APUs themselves, the reason is they lack bandwith coming from the RAM, as dedicated memory for them is pretty expensive (in form of eDRAM at least, just check Intel chips with Iris Pro graphics). If there wasn't this problem, I'm pretty sure AMD would already have released an APU which would compete at least with the XBO in performance. With Zen and DDR4-3200 this will already come close, but if AMD adds some on-chip/on-die memory (like HBM or HMC) a octacore APU with 16 Compute Units (PS4 has 18) would be very possible.

Moore's law is still in place, you just need to look in the right direction, and that's servers. But I agree to a degree, due to lack of a worthy rival for Intel (80% of the TOP500 servers have Intel CPU in them), it will probably slow down here too unless AMD and/or IBM get competitive again.

I do believe however that XBO might get a sucessor first, as it's struggling even more with 1080p/60fps than it's main rival (and saleswise too, of course), but when the XBO gets it's sucessor PS5 must follow soon to stay competitive.

crazy prediction

generations are not going to last longer than they did in the past. The reason the PS2 managed to sell well for like a decade was because when it first arrived on the scene (that whole generation with Gamecube/Xbox) it was SUCH an improvement over the prior generation in terms of technology and visuals. Also the fact that it was coupled with a DVD player

constant micro improvements to computer gaming, which already has a large lead to consoles, will just further push console manufactuers to having shorter generational period

the problem with consoles currently is that they can't easily be improved or replaced in terms of parts. what you have at the start is what you have at the end in terms of hardware. 

which is a big reason why I think in the next 10 years you will see a giant leap to home computers and consoles merging. at this point computers have a massive advantage in terms of being able to replace parts as new games and technology arrives, and that will only grow

I mean realistically in 5 years a mid tier computer will be capable of playing games visually at such a higher level than the PS4 that it won't even be funny. as it stands mid tier computers can play games much better to begin with! just realistically you can't expect consumers to consistently purchase this system that far down the road

a 10 year life cycle is unrealistic. obviously expect the PS4 to pass 100 million in the next 2 or 3 years but its peak year is probably 2016 and undoubtedly another big console (besides the NX) will be released either prior to 2020 or right around then. mark my words. its just inevitable due to computer pressure and the possibilities that exist from decent PC's. people obviously will always still buy older systems but at the rate things could change in the next 5 years its unrealistic to expect it in the numbers you're predicting

things like the PS2 and the DS are exceptions and I would argue that the current PS4 dominance is partially DUE to Microsoft and Nintendo dropping the ball a bit. for all we know the NX could be a giant hit like the Wii and something like that alone would absolutely kill the momentum of the PS4 coming close to the milestone you suggest. too many variables that point to it not happening

if you had 5 more guranteed 2015 years then, yes, it would seem possible- but there is no example of a generation in which a system could maintain that sort of level of sales for so long. unless you doubt that 2016 is likely the peak year for the PS4 (it probably is though regardless of what you think)

"I looked at trends from past consoles and I arrived at this"

That should start the alarm clocks right away, shouldn't it? This is not how it works and it's never this simple. People did the same thing for the Wii, showing how it was tracking way above the PS2 from the start and was sure to destroy its lifetime sales record, it has also been done for the 3DS and people had insane predictions for the sales of the PS3 in its twilight years, even stating that it would catch the Wii due to the infamous ten year plan.

I don't think the PS4 will even come close to 140 million, beating the PS1 would be a feat in today's market and is a more realistic goal. "At least 140 million" is way over board at best. The 7th gen was unnaturally long; it was the exception to the rule due to HD being ushered into the industry, rising development costs that required completely reworked developer paradigms, exploding online communities and features and an unprecedented close horse between the three consoles.
Also, mobile gaming, browser and social games were much less of a factor and we had a huge, global financial crisis to deal with.
Circumstances caused the 7th gen to be really long, these are not present in the 8th gen, using the 7th gen to predict the length of the 8th gen is almost as faulty as using past console sales to predict the lifetime sales of the PS4.

please don't use "LTD" as "life to death" or sth, the abbreviation already is in use and means "life to date", sales up to a specific date, better just say "lifetime"

I think it's possible for the PS4 to sell >110m, but much more than that seems quite a bit unlikely to me right now