Extreme and Mixed-Precision Quantization: From FP8 to Binary Neural Networks31 March 2026AI Accelerator Quantization FP8 INT4 Binary Neural Networks BitNet QuIP AQLM HQQ Mixed Precision LLM Optimization Model Compression GGUF KV-Cache Vision Transformer Diffusion Models Inference Optimization