what is hezium.memory fr?

Hezium.Memory is a .NET library that lets developers work with arrays far larger than the built in Array and Span types normally allow. Standard .NET arrays have a maximum size limit, and this library provides drop in replacement types called BigArray, BigMemory, BigReadOnlyMemory, BigSpan, and BigReadOnlySpan that use a wider number type for lengths and indexes so a single collection can hold billions of elements while still feeling like the familiar array and span programming style developers already know. The library is aimed at situations where you naturally have a huge amount of one dimensional data, such as columnar datasets, large lookup tables, buffers for talking to native code, generated datasets, simulation state, or files processed as one continuous block. Rather than inventing a new way of thinking about indexing, slicing, searching, copying, and sorting this data, Hezium.Memory intentionally mirrors the existing Array, Memory, and Span APIs that .NET developers already use every day, just scaled up to handle sizes those built in types cannot. Under the hood, BigArray behaves like a normal array backed collection for smaller sizes, but automatically switches to a chunked internal storage scheme once the requested length exceeds what a single managed array can hold, all while keeping one continuous index space from the caller's perspective. It also integrates with .NET's garbage collector, supporting both regular managed allocation and explicit pinned or uninitialized allocation similar to the built in GC helper methods. The package is distributed through NuGet and installed with a single command, and it is written entirely in C#. It is a low level building block intended for developers already comfortable working with spans, memory, and manual buffer management in .NET, rather than a general purpose end user application.