I really hope this performs much better than UE5. Maybe they announced it already to favor early adoption and move from UE5 as quickly as possible.

I remember Tim Sweeney did a interview last year where he talked a little about his plans for Unreal 6. From what I remember the most exciting change from a performance standpoint was making the Game Simulation Logic no longer run on a single thread. Only downside of this is it harder not easier to program game logic on multiple threads but I still think it worth it for the performance advantages, it 2026 limited anything to one thread at this point seem so dated.

Complete list of changes I remember he mentioning is

1) Make Game Simulation Multithreaded
2) Combining Unreal Editor for Fortnite into Unreal Engine
4) Adding the Verse Programming Language (currently used in unreal editor for Fortnite but not unreal engine 5)
3) Metaverse something something

UE5 always felt like it was too ambitious for console hardware; Lumen, Nanite, and VSM tended to absolutely tank performance.

Their priority for UE6 should be to get it running well on the hardware people actually have rather than shooting for the moon.

UE5.6 is supposed to be much better optimized. We'll see how well when games designed for it (like Witcher 4) release and if those improvements carry over to UE6.

At a quick glance, that actually seems like a really good language. There's some less typical parts I'd have to get more experience with to form a proper opinion, but my initial impression of the language is really good.

Hopefully; thus far the engine's been disappointing to say the least. They have been some good looking games on it like Hellblade II and Layers of Fear, and Witcher 4 does look phenomenal if what they've shown so far is legit, but so often it seems to bring with it blurry resolutions and other steep compromises.