The pdf of the paper is here
He is referring to gate model quantum computers with 1000 to 1,000,000 qubits.
Others think Dyakonov has got it wrong. "It is true that the quantum computing community should be cautiously optimistic, rather than confident," says Andrew Steane of the University of Oxford. "But his arguments are largely misleading."
He also is not talking about adiabatic quantum computers like that being worked on by Dwave.
adiabatic quantum computer