In the setting of secure multi-party computation, two or more parties with private inputs wish to compute some joint function of their inputs and achieve the security requirements of privacy, correctness, independence of inputs and more. Secure multi-party computation is not only the general basic research of secure protocol, but also applied in many applications such as coin-tossing, electronic voting and private information retrieval schemes. The research of secure multi-party computation provides a central tool in many area of cryptography. In recent years,secure multi-party computation has been advancing in leaps and bounds, especially in the practical techniques; the practical technology becomes a new attractive field in secure multi-party computation. In this paper, we introduce the main advances and results of practical secure multi-party computation, and focus on three major supporting techniques, which include garbled circuits optimization, cut-and-choose technique and oblivious transfer extension.These techniques significantly improve the efficiency of secure multi-party computation in different aspects.