Will computers ever solve chess?

Memory is not the problem: traversing every relevant possibility is. High confidence solutions are far more feasible, like probabilistic primality tests.

Chess will not be solved for commercial reason. Full Solution means death of the game and  further there will be no business issues.

Let's be frank, chess is commercially insignificant, like croquet.

In the grand scheme of things, yes. So what does that make chess? 

Whether the game is solvable is a completely different question from what the result of such a solution would be, so the thread question implies no such assumption. There is really no way of saying whether a perfect game of chess is a draw, a white win, or a black win for that matter. Of course a draw is likely, and this is the opinion of nearly all qualified players on the subject, but we really do not and cannot know until the game is actually solved, and this will be for a long long time a practical impossibility as many noted in this thread already.

On a sidenote, the concept of an ELO rating becomes meaningless with perfect play, so there is no point in speculating what that would be either... Ratings are a statistical measure that express the relative strength between two players, an expected win percentage in other words. When two players play only perfect games there is no expected outcome, all outcomes would be precisely known in advance (only white wins and black losses or the other way around, or only draws).

You are incorrect there. If two players are perfect, they have the same Elo rating. Against other players they get better results. That is what defines their rating. There is some rating for a perfect player, resulting from the totality of results in a pool including players whose rating is currently known.

As far as I understood you cannot improve your position in a deterministic game - like chess. You only can make it worse. Of course we use this term when going for a subjectively better or more stable position. However, when the position allows this setup in terms of tempi the assessment of the original position is the same as the final position.

You would never assess a K+P vs. K endgame with pawn on 2nd rank differently than the one with pawn on 7th rank. The endgame, by force, is either won or a draw and both sides can only make it worse.

"Perfect" chess is probably somewhere between Regan's 3571 calculation and Railich's 3600 estimate. 

This means that a 2800 is likely to draw against a perfect player about 1 game in every 50. That's really depressing, if accurate.

True, with only perfect players all ratings would necessarily be the same, but when non-perfect players participate in the same rating pool there would still be ELO ratings in a practical sense. But even then the rating of a perfect chess player would not be a measure of the quality of his play, as all of his moves are perfect already, i.e. always the best move possible. It would merely reflect the measure of imperfection of all other players combined, not a meaningless number, but not a measure of the perfect players strength either.

So much false in 1 paragraph :).

Computers are light years away from proving a forced draw.  We'd be very lucky to have a 10 piece tablebase in the next decade.

Video games are a horrible, horrible example.  Video games are designed to be beaten.  All of them.  Every single one.  Any game developer could write a game of pong in 10 minutes that would throw the ball so fast that you wouldn't even see it...but nobody would ever play that game.  Even "nightmare" modes are specifically designed to be beaten.  It's always possible to just crank up the difficulty past human reaction times...but nobody does that because they, you know, want to actually sell games.

I very much doubt there will ever be a human beating even todays chess engines with whatever we will learn from any source, nevermind the fact that it would be logically impossible to beat a future computer who has solved chess.

I am intrigued though: what do you mean by humans having regained supremacy over computers just like we did with every arcade video game ever created ? Could you please explain that some more ?

This is the dumbest post in chess.com. Owner of chessdotcom become millionaire by doing business on chess. Komodo writers become millionare by doing this business.

Moreover see a calculation for chess engine business:

Suppose,

Price of engine per softcopy = $50 dollar
Number of sold copy = 10000 [though it is even more]

Total revenue = $ 500000 per version

Number of versions = 10

Grand total revenue = $50,00,000 [ 5 million]

So, More version more income. If engines reach their highiest potentiality, there will be no business

But most dangerous thing is solving chess as database/tablebase. That is really very bad. Cause first it will kill the top level chess in the world. Genius & stupid will be in same row.

How? Genius and stupid would still be just as large a difference as they are today. Sure it would change a lot of opening theory and some endgame theory, middlegame plans will be finetuned etc, but in the end there are still imperfect finite human minds playing eachother over the board (and forget about remembering everything because it can't be done).

  How else would Capa justify his main pleasure? But in actuality, what is chess? Nothing but a game. And all games are stupid, intrinsically, because they are just games, no matter how sophisticated, that is the limitation of any game. Humanity is not better, nor will ever be, as a result of playing a game, any game.

  But you will never even come close to its calculating power, which is all it has, including its evaluation errors.

This isn't really true. You should put yourself in the shoes of chess engine designers.

If you are the first company to build a chess engine that solves chess completely, you are not going to lose any business.

Rather, you will now have the rights to the only chess engine that has chess completely solved and now every chess person to ever live will buy only your chess engine for ever and ever.

Expected payout to whoever solves chess: Infinite

Ok. 

I predict that the rest of the world will have an engine that never loses in about 20 years. An actual mathematical proof that chess is a draw will be many years in the future, if ever.

Of course, most GMs accepted that chess is a theoretical draw before Capablanca's days. But that is just a belief, not a proof.

Even when chess is solved, the mystery gone, the game is still exciting. No human genius have mastered the 7 men EGTB. What to speak of 32 men EGTB.

  Their results can. But to arrive at those results ( the best move) calculating 200 mil. moves per second is not exactly tic-tac-toe. That is, in open, tactical positions.

Might as well ask if computers have solved billiards. Program a robotic arm to shoot the best angles in pool. People will still prefer to play each other. Chess was also created to be played against our fellow man. Computers can't appreciate the beauty or risk of a well played match. A computer cannot solve why chess captures the mind as it does.