#Apple Silicon
11 related articles
A Practical Guide to Deploying oMLX Local Inference on M5 Max 128GB
A complete deployment and tuning guide for oMLX local inference on M5 Max 128GB: server parameters, multi-model selection, Claude Code integration, speculative decoding speedup data, and key changes in 0.6.4.
How Fast Can Macs Really Run Local AI? A Real-World Benchmark of M1 Through M6 Chips
M1 to M6 Mac chips for local AI: real benchmarks reveal pre-fill vs decode speed, memory bandwidth bottlenecks, the M5 Neural Accelerator leap, speculative decoding gains, and why RAM matters most.
Qwen 27B on M4 Max: The Truth Behind Local Model tokens/s Numbers
A developer systematically benchmarks Qwen 27B on M4 Max MacBook Pro, revealing TTFT and prompt processing realities beyond tok/s. MLX is ~4x faster than GGUF; 128K context takes 21+ minutes to prefill.
M5 Ultra AI Performance Benchmarks: The Most Powerful AI Mac with Nearly 5x Matrix Computing Performance
M5 Ultra AI benchmark: local video generation, LLM Prefill & Decode vs M3 Ultra. Matrix compute up to 4.79x faster. Deep dive into Neural Accelerator and Ultra Fusion.
M5 Ultra Mac Studio Review: A Performance Leap for Local AI Workstations
M5 Ultra Mac Studio review: 68% CPU multi-core boost, doubled ray-tracing GPU performance, and 256GB unified memory that runs DeepSeek 284B locally — the ultimate private AI workstation.
The Technical Challenges of Developing a Linux GPU Driver for the M4 Mac Mini in One Month
Developer Cody Ho built a Linux GPU driver for the M4 Mac Mini in one month. We break down the core challenges of reverse engineering Apple Silicon's closed GPU architecture.
Running Claude Code Locally: A Fully Offline AI Coding Solution for Apple Silicon
claude-code-local runs Claude Code 100% offline on Apple Silicon via MLX, supporting Qwen 3.5 122B at 65 tok/s — built for NDA, legal, and medical privacy workflows.
TutorialsoMLX + MTP + Qwen3.6: Local AI Coding Speed Breaks New Records
Using oMLX with MTP and Qwen3.6 35B on Apple Silicon Mac to achieve 86.7 tokens/s local coding speed, building a full-stack app in under 5 minutes.
Tech FrontiersMay 2026 Apple Developer Newsletter: Accessibility Labels and Silicon Migration Essentials
Analysis of Apple's May 2026 Hello Developer newsletter covering Accessibility Nutrition Labels, Apple Silicon migration, Infinity Nikki interview, and developer action items.
Industry InsightsApple's Internal Design & Development Workstation Configuration Revealed: The Product Philosophy Behind Standardization
Apple equips internal design and development teams with standardized workstations. Analyzing the logic behind Apple's unified dev environment, comparing Silicon Valley workstation strategies, and exploring Apple Silicon's advantages.
Product ReviewsSafari MCP: The Most Resource-Efficient AI Browser Automation Solution on Mac
Safari MCP natively drives Safari via AppleScript, providing AI agents with 80 automation tools. Compared to Chrome DevTools, it reduces CPU usage by 40-60% on Apple Silicon, preserves login state, and runs silently in the background.