You could possibly overlock your CPU. I wouldn't recommend this on a laptop however since there's a lack of airflow and you can't upgrade the fan/heatsink (in all cases? most?). I also wouldn't do this if it's a work laptop as there's potential to damage components. If it's not a work laptop, why the fuck are you using your own resources for work anyway?
Can someone optimize the software itself? I'd look there first; something as simple as caching intermediary results can have a HUGE impact on speed. There may be some easy gains there. There may be some slightly more involved ones. This may be helpful: http://www.mathworks.com/products/parallel-computing/parallel-computing-on-the-cloud/
Another possible avenue may be to just automate the whole thing. If you can distill the simulation tweaking, given the results, into a set of discrete deterministic steps, then you can write a program to do the tweaking for you and rerun the simulation. Then you can just let it run 24/7 until you get what you need.
More RAM is useless unless you're experiencing disk thrashing; hard to say if this is an issue without more information... like how data is being read/stored (and from what), the processes working set size,
and how much RAM you have.
SSD will also not be helpful at all unless you're bottlenecking on reading/writing data. Need more info.
There's little shit you can do like increasing the priority of the process, and disabling non-essential services/processes but .. it's unlikely to have much of an effect. The biggest thing would be to disable any anti-virus software real time protection; get permission from your work if you do this.