I don't think that there is a way to do this on one laptop.
Do you believe that you need to synchronize the sampling for each IMU to the same clock?
Do you have specific performance criteria that you are trying to achieve or is the purpose of your experiment to see if you can improve the noise by sqrt(2)?