After two major nuclear power plant accidents in Chernobyl (1986) and Fukushima (2011), one of the main requirements for the nuclear power engineering is the safety of the nuclear reactors in operation, as well as new nuclear power plants of the fourth generation, which are being developed now. One of such requirements is presence of the so-called “inherent safety” mechanism, which renders the uncontrolled reactor runaway impossible under any conditions, moreover, the implementation of such a mechanism should be ensured on the level of physical principles embedded in the reactor design. Another important problem of the nuclear power engineering is the need of the transition to the large-scale use of the fast-neutron breeder reactors, with which it would be possible to set up expanded reproduction of the nuclear fuel and by that means solve the problem of supplying humanity with relatively cheap energy for thousands of years. Moreover, at present an unresolved problem is the disposal of spent nuclear fuel containing radioactive nuclides with long half-lives, which presents a long-term danger to the ecology. One of the promising conceptions of the fast-neutron breeder reactor, which can, in the case of successful implementation, partially or even entirely solve the problems of the nuclear power engineering mentioned above, is the reactor that operates in the nuclear burning wave mode, which is also known as “Traveling wave reactor”, CANDLE and by some other names. This paper presents a short review of the main theoretical approaches used for description of such a physical phenomenon as slow nuclear burning (deflagration) wave in the neutron multiplication medium initially composed of the fertile material 238U or 232Th. A comparative analysis of the possibilities of different mathematical models for describing this phenomenon is performed, both for those based on the deterministic approach (i.e. solving neutron transport equations) and for models that use Monte Carlo methods. The main merits of the fast breeder reactor, working in the nuclear burning wave mode, and problems related to the practical realization of the considered concept are discussed.