[Unofficial] Machine solves cube in 4.9s

Ooh cool.
Good idea having the turning mechanism physically attached to the centers.
I wonder if one of the reasons it's a little quicker than others is because when it does the scramble it knows what the position of the cube is, rather than having to take images of the different sides and determining the state.
Hmmm, I wonder if corner cutting could be incorporated into the turning to speed it up a little.

Is that CFOP I see? ._.

Ah, the move count is quite high for the solve isn't it. Also the scramble sizes are quite small (12 and 14 - that's a little low for a random state isn't it?)
 
Last edited:
probably because roux is slow when you don't have M slices

BURN


...but yeah, you should really use random-state scrambles and a better solving method. Unless the point is to demonstrate the limits of a CFOP solve, in which case go ahead, I guess.
 
Try it yourself - see if you can figure out a way to hold the cube totally steady, and do 10+ tps for any move sequence, while allowing all layers to potentially rotate. It's not easy... far easier to just hold the centers constant.
 
Last edited:
There are five axeses only. No green, defined as D. With CFOP, D is not needed.
I'm looking for a better means than CFOP. But there should be no D and no M. M can be replaced LR' but M might create D, even there is no D.
Who can help me?
 
Back
Top