PixiJS sprite bounding performance degrades with large datasets

When rendering a large number of sprites in PixiJS, calling getBounds() on each sprite or on the container becomes expensive, leading to frame‑rate drops. Developers often paginate the dataset by slicing the sprite array or swapping texture atlases, but it is unclear whether fetching data in chunks from a REST endpoint or handling pagination entirely on the client side yields better overall performance.

The goal is to determine which pagination strategy maintains acceptable FPS while keeping memory usage within practical limits for typical web‑application scenarios.

• What are the performance trade‑offs between server‑side REST pagination and client‑side chunking with texture atlases?
• How does each approach affect memory consumption and latency during sprite updates?
• At what approximate sprite count does one strategy begin to outperform the other?