Generational GC, large object heap, boxing costs, spans, and diagnosing memory pressure.
3 questions in .NET & C#.
The GC segregates the heap into generations on the assumption that most objects die young, collecting gen0 often and gen2 rarely, with large objects on a separate heap that is not compacted by default. Allocating less means removing boxing and using Span<T> over pooled or stack memory.
The kernel kills you on everything in the cgroup, while the GC only reports the managed heap. Committed-but-empty heap, per-core Server GC heaps, thread stacks, native buffers and loaded code all count against the limit and none of them appear in a heap size figure.
Span is a stack-only view over memory you already have, so slicing it costs nothing while Substring allocates a new string and copies. It removes allocations rather than making them faster, which is the right lever in a runtime where the cost is collection pressure rather than the allocation itself.