Systems in motion

Three ways to test what you think you saw

This issue connects three ideas: a source can settle a factual question, a frame-time trace can expose what an average hides, and changing measurement basis can reveal a difference that was invisible before.

A city is a scheduling problem

An open-world engine has to deliver assets, simulation and rendering work on time as the player moves. The technical feature explains how those jobs overlap and what Rockstar’s PS5 capture statement can establish.

Look past average FPS

Two synthetic traces each deliver 62.5 fps on average. Only one contains five 54 ms intervals. Run the Lab script to inspect the percentile and frame-budget numbers yourself.

Change the measurement basis

Two Bell states look identical in one measurement basis and disagree in another. The Python simulator turns that difference into a repeatable four-number result.

• GTA VI engineering analysis
• Frame pacing experiment
• Bell-state simulator