Two renderers share one L40S: voxlap ported to CUDA Rust streams multiplayer views in real time
idanbeck · x · 2026-10-03
Developer idanbeck ported the classic voxel engine voxlap to CUDA + Rust and ran it on the inference stack he's been building, achieving near-real-time streaming from a single L40S GPU.
He then added multiplayer via frame "batching": with a shared environment, one GPU renders from multiple viewpoints and streams to several clients at once — the demo shows 2 renderers on one GPU, and he believes it can scale to 16 or even 32.
The unexpected result: running a legacy 3D engine on a modern GPU inference pipeline works surprisingly well, with a fuller demo video promised.
More from Infra
- Fireworks launches cache-aware FireRouter with Opus: coding costs cut 57% at 98% accuracy — Madisonkanna · 2026-10-03
- OpenAI reportedly weighed $100M Hugging Face investment before Nvidia deal, with chip distribution in play — VraserX · 2026-10-03
- Strata on a single RTX 3090: 256k context at 38-61 t/s, 2x faster than llama.cpp — cezarducatti · 2026-10-03
- "Opposing cheaper electricity because data centers benefit" sparks AI power-policy fight — 2C_ornot2C · 2026-10-03
- Jev Decision Models Cut Edge Orchestration Latency 22.7-64.5% vs LLMs — Delong Li · 2026-10-03
- Anthropic to spend at least $518B on AI infrastructure over a decade, IPO filing shows — Beth_Kindig · 2026-10-03