Hi, I optimized a world model, Lingbot-World 2.0 1.3B, to run with real-time 16ps on an RTX 5090. It's 2.7x faster than Robbyant (Alibaba's lab), 2.5x vs SGlang, and 1.9x vs Nvidia FlashDreams engine. The drawback is that the resolution is 832×464, so you'd have to play it with a small window. It maintains lossless performance while running on a 1x consumer GeForce GPU for a model that's very compute-bound and batch size = 1! The majority of the wins came from:
- running the decoder in half-precision with fp16 instead of fp32
Very interesting, congrats, and thanks for sharing. What are the major advantages to running something like this as opposed to running a simple Unreal Engine scene?
Hi, I optimized a world model, Lingbot-World 2.0 1.3B, to run with real-time 16ps on an RTX 5090. It's 2.7x faster than Robbyant (Alibaba's lab), 2.5x vs SGlang, and 1.9x vs Nvidia FlashDreams engine. The drawback is that the resolution is 832×464, so you'd have to play it with a small window. It maintains lossless performance while running on a 1x consumer GeForce GPU for a model that's very compute-bound and batch size = 1! The majority of the wins came from:
- running the decoder in half-precision with fp16 instead of fp32
- switching FlashAttention to SageAttention
- writing some custom kernels (minor)
Currently works on Linux only.
Run: $ lingbot play dragon
Very interesting, congrats, and thanks for sharing. What are the major advantages to running something like this as opposed to running a simple Unreal Engine scene?