The synchronization control problem of memristive neural networks in complex dynamic environments was investigated. A preset-time synchronization strategy was proposed based on Lyapunov analysis. Finite-time synchronization was guaranteed by the proposed method. The convergence time could be predetermined according to system requirements. Therefore, flexibility and efficiency were achieved in theoretical design and practical implementation. The development background and applications of memristive neural networks in intelligent computation were reviewed. The limitations of conventional synchronization strategies were analyzed. A new controller design framework was then formulated. Sufficient conditions for the effectiveness of the method were established through theoretical derivations and stability proofs. Comparative simulations between direct preset-time (Direct-PT) control and time-space deformation (TSD)-based control were conducted. The results demonstrate that the proposed strategy has clear advantages in convergence speed, robustness and disturbance rejection capability.