How does .populate() work when many documents refer to the same model?

[nikzanda]

If many documents referring to the same model are involved in a query, does .populate() execute one query per ObjectId, ignoring duplicates, or does it execute only one query grouping all ObjectIds?

Example:
I'm trying to understand which approach offers better performance when dealing with many students referencing the same schools.

Option 1:
First, I query all students. Then, I extract unique school IDs and query the School collection separately to build a map of schools. Finally, I iterate over the students and retrieve each student's school from this map manually.

const students = await Student.find();
const schoolIds = [...new Set(students.map((s) => s.school.toString()))];
const schoolsMap = await School.find({ _id: { $in: schoolIds } }).transform(/* array of schools to Map<string, School> */);
students.forEach((student) => {
  const school = schoolsMap.get(student.school.toString())!;
  // ...
});

Option 2:
I simply query all students and populate the school field directly, then access student.school during iteration.

const students = await Student.find().populate('school');
students.forEach((student) => {
  const { school } = student;
  // ...
});

My question is:
Which approach generally provides better performance, especially when many students refer to the same schools? Are there any differences in how .populate() handles the underlying queries in these two scenarios?

[vkarpov15]

As written, there shouldn't be any difference. Student.find().populate('school') executes 2 queries: 1 to find all students, and 1 to find all corresponding schools using { _id: { $in } } with deduplicated school ids, just like Option 1. Performance difference should be minimal, if any.