1: \begin{abstract}

2: In this paper we are interested in the numerical solution of

3: stochastic differential equations with non negative solutions. Our

4: goal is to construct explicit numerical schemes that preserve

5: positivity, even for super linear stochastic differential equations.

6: It is well known that the usual Euler scheme diverges on super linear problems and

7: the Tamed-Euler method does not preserve positivity.  In that

8: direction, we use the Semi-Discrete method that the first author has proposed in

9: two previous papers. We propose a new

10: numerical scheme for a class of stochastic differential equations

11: which are super linear with non negative solution. In this class of stochastic differential equations belongs the Heston $3/2$-model that appears in financial

12: mathematics, for which we prove %theoretically and 

13: through numerical experiments the ``optimal'' order of strong convergence at least $1/2$ of the Semi-Discrete method.

14: \end{abstract}

15: