Zyphra's latest release shows that an autoregressive MoE model can be converted into a discrete diffusion model with no systematic loss in evaluation performance. ZAYA1-8B-Diffusion-Preview achieves up to 7.7x inference speedup over autoregression by shifting decoding from memory-bandwidth bound to compute-bound — a key advantage as modern GPUs continue scaling FLOPs faster than memory bandwidth. The post Zyphra Releases ZAYA1-8B-Diffusion-Preview: The First MoE Diffusion Model Converted From an Autoregressive LLM With Up to 7.7x Speedup appeared first on MarkTechPost.