Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: speculative-decoding
BitNet and Cassandra: How Ternary Weights Speed Up Local LLMs on CPUs
Six findings on 1.58-bit BitNet, speculative decoding and Cassandra: local LLM inference up to 6.17x faster on x86 CPUs, 82.2% less energy, same accuracy.
4 min readOpenVINO 2026.4: MTP, DFlash, Node.js ASR and New Models
What's new in OpenVINO 2026.4: MTP speculative decoding, DFlash for Qwen on Intel GPUs, ASRPipeline in Node.js, and a long list of CPU, GPU and NPU models.
5 min readDoubling Qwen3.8-Flash-Next Decode Speed with MTP on RTX GPUs
Learn how to build a llama.cpp fork with MTP speculative decoding and the right flags to roughly double Qwen3.8-Flash-Next generation speed on multi-GPU RTX rigs.
5 min readSpeculative Decoding, Early Exit and Quantization: Faster LLM Decode
How speculative decoding and early exit cut LLM inference latency, when draft acceptance pays off, and how they stack with quantization and KV cache paging.
8 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.