[en] A possible approach to choosing the power level for an experimental reactor, based on correlation between the reactor power and the volume of the information obtained is discussed. Calculation results for the experimental reactor with the 1x10¹⁶ neutron/cm²xS flux density are presented to illustrate the approach proposed. The main conclusions are the following: with the reactor power increase in the range up to approximately 400 MW the total information volume increases in an approximate proportion to the cube of the power increase, and specific information and the rate of information obtained increases in an approximate proportion to the square of the power increase. With the power increase above approximately 400 MW the growth of the total information volume slows down. In a large reactor the same level of neutron flux density can be attained at significantly lower power intensity of the core, that facilitates the problem of fuel element construction for a high-flux reactor. The great power of the experimental reactor permits to use fuel of a relatively low inrichment. It means that a rather high fuel conversion coefficient (approximately 0.5) can be obtained that to a certain extent facilitates the problem solution of reactivity compensation in the experimental reactor