Convergence improvement

Is there a particular improvement added to QSPICE with regards to convergence when compared to LTSPICE? For example, I use some op-amp models and with LTSPICE it takes a lot of time to find the DC operating point. In fact, it fails to obtain it, so much that I had to use .options noopiter and GminSteps = 0 to avoid having to wait for ages. In QSPICE there is no such problem and the simulation is almost instantaneous using the same op-amp model.