qwenGPU compiles a 4-lane SIMT GPU into Qwen3.5-2B as 45 untrained layers. The model writes every step of real CUDA launches itself, using Gated DeltaNet state as registers and memory.
This article has been indexed from MarkTechPost
Read the original article:
Meet qwenGPU: Qwen3.5-2B With 45 Compiled GPU Layers That Run CUDA