Proper subgroups of non-cyclic p-group cannot be all cyclic?

Prove or disprove?

I'm leaning towards it being true but not sure. Any hint would be greatly appreciated.

In case of it being false, i.e a non-cyclic p-group can have all cyclic proper subgroups, is there any way to count them?

-

If you take the smallest non-cyclic $p$-group, then its proper subgroups are smaller $p$-groups and thus have to be cyclic. So, it can happen.

-
And if it is not the smallest non-cyclic p-group, i.e if I have already proven there is one of smaller order? –  ikk Nov 4 '12 at 10:09
@ikk Do you understand that a single counterexample disproves the statement in you title? As far as your new question goes, I suggest that you identify the smallest non-cyclic $p$-group and check whether it is a subgroup of all other non-cyclic $p$-groups. –  Phira Nov 4 '12 at 10:17

Take $V_4$, the Klein group of order 4 or the quaternion group $Q$ of order 8, or the dihedral group $D_4$ of order 8. Those are the smallest examples of non-cyclic groups with only proper cyclic subgroups.

-
Erm - isn't the Klein 4-group the smallest non-cyclic group with all of its proper subgroups cyclic? –  Old John Nov 4 '12 at 15:25
Yes, edited. My computer started to reboot just before I wanted to edit this. Thanks. Back online again. –  Nicky Hekster Nov 4 '12 at 15:37