1: \begin{abstract}

2: The recent advance of algorithms for nonlinear semi-definite optimization problems, called NSDPs, is remarkable.

3: Yamashita et al. first proposed a primal-dual interior point method (PDIPM) for solving NSDPs using the family of Monteiro-Zhang (MZ) search directions.

4: %The MZ family was originally proposed in the context of linear semi-definite optimization problems (LSDPs).

5: Since then, various kinds of PDIPMs have been proposed for NSDPs, but, as far as we know, all of them are based on the MZ family.

6: In this paper, we present a PDIPM equipped with the family of Monteiro-Tsuchiya (MT) directions, which were originally devised for solving linear semi-definite optimization problems as were the MZ family. We further prove local superlinear convergence to a Karush-Kuhn-Tucker point of the NSDP in the presence of certain general assumptions on scaling matrices, which are used in producing the MT scaling directions.

7: \end{abstract}