DataLens is an engine-agnostic, cache-aware, in-memory simulation database designed to handle large-scale, emergent actor systems encompassing hundreds of thousands of concurrent entities. Built on a strict three-layer architecture, it decouples passive, ultra-low-level data storage from volatile engine-specific presentation loops. By organizing all simulation state into tightly packed, cache-line-isolated, column-major arrays, DataLens completely eliminates pointer-chasing, runtime heap allocations, and heavy relational lookups from the execution critical path.

At its core, DataLens introduces the concept of the Lens: a highly optimized gameplay frustum that materializes a small, hydrated, row-major viewport of only the actors currently in active gameplay scope. While ordinary gameplay code operates safely on these isolated, double-buffered snapshot views without the risk of data races, background Systems execute directly over the flat data columns, chunking workload across threads and utilizing branchless vector kernels to process massive entity loads at variable simulation cadences. The result is a hyper-fast bit-blitting substrate that serves as the deterministic, memory-safe foundation for complex state engines and systemic simulations alike.

As with all Heathen tools DataLens is available for multiple engines:

And is composed of two packages: