<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/bb99b751f1934558abc33971ee535244&quot; frameborder=&quot;0&quot; width=&quot;2060&quot; height=&quot;1545&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1545</height><width>2060</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1545</thumbnail_height><thumbnail_width>2060</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/bb99b751f1934558abc33971ee535244-6b10d5d4181cbb58.gif</thumbnail_url><duration>1229.785</duration><title>Parallel Execution, Scheduling, and Benchmark Discipline</title><description>Cruz introduces the Parallel REVM Lab, a public case study on deterministic EVM state machine execution and base storage contention. He explains the lab is not a production execution client and does not claim real TPS or gigagas per second speedups, focusing instead on the engineering process. The main goal is showing how to decide what can safely run in parallel versus what must serialize, with emphasis on concurrency trade-offs, scheduler correctness, and benchmark discipline. He frames the work as a broader approach to deterministic state machine execution beyond EVM specifically.</description></oembed>