#quantization

• [0007]2026-05-19Unleashing Blackwell's 4-bit: a surgical look at MXFP4 and NVFP4#cuda#quantization#fp4#nvfp4#mxfp4#blackwell#sm120#rtx#5080
• [0006]2026-05-13From 429 GB/s to the DRAM wall: writing an FP8 quantizer on an RTX 5080#cuda#quantization#fp8#blackwell#sm120#rtx#5080
• [0005]2025-10-18W8A16 Quantization with LLM.int8-Style Outlier Handling#quantization#int8#transformers
• [0004]2025-09-07How Critical Are Outliers in Transformer Models? A Live Experiment - Phase 1#quantization#transformers#experiment

all posts →