Hi aokada09,
It looks like you are running into issues that arise from the parallel execution of the two loops you have. There isn't anything necessarily tying the two loops together, so they each iterate at the rates you specified but are allowed to start whenever software (sort of randomly) dictates that they start.
To get solid synchronization, share a sample clock between the two measurements but run the sample clock at 100Hz for the pressure measurement and 50 Hz for the thermocouple measurement. You should be able to use the shared clock in the backplane of the PXIe chassis. This will be as close as possible to synchronized without the use of an additional timing card. The only real source of delay/skew between readings at that point is the physical distance that the clock signal must travel; it will not result in large skew, but there will definitely be some (probably on the scale of high-nanoseconds or low-microseconds). This article gives more information about synchronization and sample clocking:
http://www.ni.com/white-paper/11369/en/
Matt | NI Systems Engineering