Inefficient broadcasting in pytorch

I made a broadcasting but this consumes extremely high resources on my computer that even when running it in the cloud the server restarts for lack of resources, and my computer froze several times when trying to run the code. The broadcasting I did in pytorch in python, this is to compare whether it is true or false a row tensor with a column tensor. for example entry_time_boolean = entry_time <= date . where entry_time has 1 dimension and date has 2 dimensions

What are the sizes of the used tensor and their dtype?