How many Qubits need to slove chess?

[jp]

Hmm, maybe it is not so much about new algorithms, but how to implement the old ones on quantum machines?

Well, yes, it would be very nice to have a large general quantum machine.

I admit I have some ideas in my drawer for an chess/game playing quantum machine, project Iota, but my
drawer works in FIFO matter, so it will take some time....

If that could be done, it would be a huge surprise. A dozen qubits is plenty to cover every possible chess position. The problem is what you do with those qubits.

[Dann Corbit]

No number of qubits will help solve chess unless there is a chess algorithm you can run to use those qubits.

There is currently no such algorithm, and there are no prospects for one.

There is a search algorithm that can run on quantum computers.
I do not know if it can be adapted for chess, but I would not rule it out.
https://en.wikipedia.org/wiki/Grover's_algorithm
It is not inconceivable that a chess algorithm for such a machine can be invented.

[Raphexon]

We already have alpha-beta pruning, chess is solved. Now we only need time.

[jp]

We already have alpha-beta pruning, chess is solved. Now we only need time.

That's not solving unless the pruning is perfect, which it cannot be unless you've solved chess by other means.

[jp]

There is a search algorithm that can run on quantum computers.
I do not know if it can be adapted for chess, but I would not rule it out.

No, it can't be. That's not like a chess-tree search.

[Raphexon]

We already have alpha-beta pruning, chess is solved. Now we only need time.

That's not solving unless the pruning is perfect, which it cannot be unless you've solved chess by other means.

A/B converts to minimax.
Just needs some time.

[jp]

A/B converts to minimax.
Just needs some time.

By "solve", we do mean both perfectly and in the lifetime of the universe. (There are more formal ways of putting it in computer science.)

[Kotlov]

No algorithm or no hardware?

[jp]

No algorithm or no hardware?

No known algorithm. Hardware with very small numbers of qubits exists.

[Raphexon]

A/B converts to minimax.
Just needs some time.

By "solve", we do mean both perfectly and in the lifetime of the universe. (There are more formal ways of putting it in computer science.)

What do you mean with perfectly?
32 piece tablebase kind of perfect?

Or weakly solved like checkers? In that it only has the perfect solution from start (initial position) and only plays perfectly if opponent also plays perfectly. (minimax)

The first would take more time than available.
The second would just take some time.