Neuronal Model Speed _ Dynamic System vs SIMULINK vs ode23

Hi all,

I built a small network of 5 neuron cells in MATLAB using ode23 as my differential equation solvers. I tried my best to optimize it so it runs roughly 17-20 seconds (I think there are roughly 100 equations in there, but I basically used as much vectorization as possible so that I wouldn't need to write each equation serialy).

But I have read that you can also do it with SIMULINK for a Hodgkin-Huxley model. I have never used SIMULINK before but I can try to learn that. But my question is: Does it run faster as a SIMULINK-component driven model or is using differential equations actually faster?

Another thing I found out was something called Dynamic System Modelling (https://www.mathworks.com/help/control/ug/dynamic-system-models.html) with transfer functions, feedbacks and stuff. This is another completely new thing to me. So I have 2 questions:

(1) Has anyone tried to recreate a Hodgkin-Huxley model in term of this concept? (2) Does this perform faster than ode23 and differential equations form?

Thank you!