#PTQ
181 related articles
Tutorials·4 min
NVIDIA Model Optimizer Post-Training Quantization (PTQ) Practical Guide
A deep dive into NVIDIA Model Optimizer's PTQ workflow, covering INT8/INT4 quantization principles, calibration methods, RTX GPU optimization, and best practices for deploying quantized LLMs on consumer GPUs.
Read more →
·3 min
Unsloth Teams Up with Thinking Machines: Dynamic 1-bit Quantization Achieves 86% Compression, Running Multimodal LLMs in 270GB
Unsloth and Thinking Machines release dynamic 1-bit GGUF quantization for Inkling, compressing the model from 1.9TB to 270GB (86% reduction) while retaining 74.2% accuracy and adding vision/audio multimodal support.
Read more →
·3 min
Running a Local AI Voice Assistant on a $50 Arduino Board: A Practical Guide to Edge AI
How to run a fully local AI voice agent on a $50 Arduino Uno Q board, covering speech recognition, intent understanding, and TTS implementation for edge AI applications.
Read more →
·3 min
Maple-Preview: 20B Ternary MoE Model Achieves 120 tok/s On-Device Inference on iPhone
Maple-Preview achieves 120 tok/s inference of a 20B ternary MoE model on iPhone. We analyze ternary quantization, MoE sparse activation, and on-device inference challenges.
Read more →
·3 min
Quantprobe: An Open-Source Tool for Running 30B LLMs on 6GB GPUs
Quantprobe is an open-source memory optimization framework that enables 30B LLMs to run at 22 tokens/s on 6GB GPUs through per-layer quantization and intelligent CPU/GPU splitting.
Read more →
·2 min
Qwen3.8-Max Preview Continues Iterating with Major Improvements in Frontend Development Capabilities
Alibaba's Qwen3.8-Max-Preview iterates daily with significant frontend development improvements. The team uses an open preview strategy to collect community feedback, promising open-weight release.
Read more →
·3 min
Comparing China's Four Major AI Labs: Strategic Analysis of Qwen, DeepSeek, Moonshot, and Ling
An insider's analysis of China's four AI labs — Qwen, DeepSeek, Moonshot, and Ling — revealing their distinct strategic bets on distribution, architecture, long-termism, and serving cost.
Read more →
·3 min
Nightcrawler: Deep Dive into an AI Penetration Testing Agent Running Locally on Smartphones
Deep analysis of Nightcrawler, an AI penetration testing agent running entirely on smartphones. Exploring how on-device AI empowers cybersecurity testing, its architecture, use cases, and risks.
Read more →
·2 min
antirez Strikes Again: ds4, a Pure C Local Inference Engine for DeepSeek 4
Redis creator antirez open-sources ds4, a pure C local inference engine for DeepSeek 4 Flash and PRO with native Metal, CUDA, and ROCm support, earning nearly 20K GitHub stars.
Read more →
·3 min
Token Economics: The New Unit of Value Measurement in the AI Era
Deep dive into how Tokens evolved from a technical concept in LLMs to the core unit of measurement in the AI economy. Exploring Token consumption explosion, cost optimization, and Token economics.
Read more →
·2 min
MiniMax H3 Local AI Video Generation: Deployment Methods, Hardware Requirements & Hands-On Analysis
In-depth analysis of MiniMax H3 local video generation capabilities, exploring hardware requirements, advantages, challenges, and the trend of AI video moving from cloud to local deployment.
Read more →
·3 min
Persistent State Machines: How INT4 Memory Cells Reshape LLM Attention Mechanisms
Exploring how persistent state machines with INT4-quantized memory cells reshape LLM attention, breaking KV Cache memory bottlenecks for long-context inference on edge devices and high-concurrency scenarios.
Read more →
·3 min
Dual RTX 3060 Running DeepSeek V4 Flash: IQ2_M Quantized Inference Benchmarked at 3.5 tok/s
Benchmarking DeepSeek V4 Flash on dual RTX 3060 GPUs with 96GB RAM at IQ2_M quantization achieving 3.5 tokens/sec. Covers hardware choices, 2-bit quantization techniques, and local LLM deployment optimization.
Read more →
·2 min
Running Local LLMs on a 24GB Mac Mini: Model Selection & Speed Optimization Guide
24GB Mac Mini too slow for local LLMs? Learn why 14B models struggle, get 3B-8B model recommendations for Home Assistant, and discover Ollama speed optimization tips.
Read more →
·3 min
LLM Post-Training on an 8GB GPU: A Practical Guide to SFT, DPO, and GRPO
Learn how to complete LLM post-training on a consumer GPU with just 8GB VRAM, covering SFT, DPO, and GRPO methods using LoRA quantization and other techniques.
Read more →
·3 min
DeepSeek-V4-Flash Local Deployment Test: $8,000 Hardware Approaches Frontier Model Performance
DeepSeek-V4-Flash-0731 scores 50 on the Intelligence Index, nearly matching the frontier model score of 51 from five months prior. We analyze local deployment, hardware requirements, and implications.
Read more →
·3 min
How Cheap Can It Be to Serve a 2.8 Trillion Parameter Model? A Deep Dive into Inference Economics
Deep analysis of the real cost of serving a 2.8 trillion parameter model. From MoE sparse activation to batching scale effects and inference optimization, revealing why model size and serving cost are less correlated than assumed.
Read more →
·3 min
Practical Guide to Setting Up a Local AI Coding Environment on MacBook Pro M4
Complete guide to setting up a local AI coding environment on MacBook Pro M4, covering Ollama, MLX, Continue, Qwen3-Coder 30B configuration, and performance optimization strategies for 32GB RAM.
Read more →
·3 min
GPT-5.6 Luna Review: Outperforms Google's Flagship at a Lower Price
GPT-5.6 Luna tops Google's flagship on the Artificial Analysis Intelligence Index while priced below Google's entry-level model. A deep dive into what this performance-cost breakthrough means.
Read more →
·2 min
DeepSeek V4 Flash Weights Open-Sourced: A New Option for Lightweight LLM Deployment
DeepSeek V4 Flash model weights reportedly open-sourced. This article analyzes its lightweight positioning, open-weight value, comparisons with closed-source models, and deployment guidance.
Read more →