Perfect World Games Details NTE City Streaming at Unreal Fest Tokyo 2026
At Unreal Fest Tokyo 2026 on November 3, Perfect World Games engineers Shi Jinwei and Zheng Shengwen explained how the urban open-world RPG NTE: Neverness to Everness streams its city without loading screens on PC, Mac, PS5, iOS and Android.
1 source 11m ago
Reporting from 1 source: 4Gamer.net.
Why it matters
The talk frames seamless open-world traversal as a scheduling and memory-budget problem rather than a faster loader, which is the constraint that decides whether a city this size holds up on phones as well as PC.
Key Facts
- Perfect World Games gave the talk "NTE: City-Scale Multiplatform Optimization" at Unreal Fest Tokyo 2026 on November 3, 2026.
- The speakers were graphics engineers Shi Jinwei and Zheng Shengwen.
- NTE: Neverness to Everness is listed for PC, Mac, PS5, iOS and Android.
- NTE places a temporary World Partition Streaming Source at the destination coordinates to begin loading before the player moves.
- The readiness check is divided into three stages: resources, world and gameplay.
Shi described teleport as a state machine that preloads the destination while the player stands still, then checks resources, world and gameplay before returning control.
Shi Jinwei put the CPU side first: the work is allocating and managing the time each process gets. On the GPU side, the graphics pipeline has to be assembled well. Zheng Shengwen handled the second half of the session, covering rendering features and per-platform optimization.
The loading range is where the memory budget bites. Because the destination loads while the origin is still held, both areas occupy memory at once, and the farther the destination the less data can be shared. Unreal Engine's Fast Geometry lowers the cost of the area itself, but Shi said it only levels out load peaks and is not the teleport control.
Written with AI by Yomimono from the 1 cited source below, including Japanese-language reporting where cited. See our editorial policy.
Add Yomimono as a preferred source on Google