power query editor slow loading

Manasaanalytics 

Real fix at scale: push transformations to the source SQL view instead of doing them in Power Query.

Keyword: huge data size. Unless transformations can be folded back to the source, Power Query processes them in memory, and as data volume increases, performance becomes progressively slower. For example, when Power Query filters in-memory, it first loads the full dataset into RAM (instead of pushing the filter to the source), then applies the filter using its local M engine row by row. That means every record is evaluated inside memory. As a result, performance depends entirely on available memory and processing power, and large datasets can become slow or even fail if they exceed memory limits. Also it matters where the source is from. It generally is faster if the data is locally stored in the device instead of being sourced remotely.