Unreal's Ex-Evangelist Is Building a New Engine - I'm Rooting For It and Skeptical

Every time a big new PC game stutters its way through the first hour while it compiles shaders, I think the same thing: the engine monoculture is showing its teeth. So when the guy who spent a decade selling the rest of us on Unreal Engine walks out the door to build something new, I pay attention.

Sjoerd De Jong - Epic's longtime "lead evangelist" for Unreal, a name plenty of you know from tutorials and level-design talks - has left Epic after roughly 12 years. His new project: a from-scratch game engine called Immens, built with a small crew that reportedly includes former Epic technical director Arjan Brussee. De Jong's pitch is blunt: Unreal and Unity have gotten so dominant that development itself has stalled into "stagnation," and the industry is at a "pivotal point."

Why an insider saying this matters

I'll be honest, "the industry needs a new engine" is the kind of line I usually roll my eyes at. Everyone thinks they can build a better mousetrap until they hit the ten-thousandth edge case in a renderer. But this isn't a hype-cycle founder who watched one GDC talk. De Jong has been elbow-deep in Unreal since he was 15, modding the 1998 original - 27 years of shipping with the thing he now says isn't enough. When the evangelist stops evangelizing, that's data.

And the symptom he's pointing at is real. Look at the review threads for almost any UE5 release over the past two years: traversal stutter, shader-compilation hitching, frametime spikes that no amount of GPU brute force smooths out. I've capped framerates, pre-compiled caches, tuned SSD headroom - the tricks help, but you can't config your way out of a problem baked into how a shared engine streams and compiles. When one renderer powers half the AAA slate, its weaknesses become everyone's weaknesses. That's the monoculture tax.

The skeptic in me isn't sold yet

Here's the cold part. A new engine is a decade-long bet, and the graveyard is full of "Unreal killers." The moat isn't the renderer - it's the tooling, the documentation, the twenty years of Stack Overflow answers and asset-store plugins that make a studio pick Unreal even when it stutters. Immens starts at zero on all of that. De Jong knows the tooling side better than almost anyone alive, which is the one reason I'd give this a real shot instead of none.

What I actually want isn't for Immens to "win." I want it to exist and be good enough to scare the incumbents. Competition is the only thing that reliably fixes stagnation. If a credible third option forces Epic to finally treat traversal stutter and shader comp as the ship-blocking bugs they are, then this project wins even if Immens itself never powers a hit.

So I'm watching this one with both hands - one clapping, one holding my frametime graph. Prove me wrong, Sjoerd. Please.

Image: DX Studio / CC BY 3.0, source: https://commons.wikimedia.org/wiki/File:Editor_full.jpg