Tag
Tinkerbox 是一个可将 LiquidAI d1-3B 等决策模型高效运行在 Apple Neural Engine 上的应用，支持通过 macOS Shortcuts 触发器实现无代码自动化，延迟低至约 40ms，并计划集成 vLLM/Qwen3.5 系列模型。
作者将 Ternary Bonsai 2 27B 量化模型迁移到 Apple Neural Engine，在 16GB M6 Mac mini 上实现约 52 tokens/s 的代码生成速度，整机功耗仅约 12W，显著优于 GPU 方案。项目基于 ANEMLL Forge 提供转换脚本、运行时补丁和 OpenAI 兼容接口，并已适配 Grok Build。
The article retrospectively reverse-engineers Apple's Neural Engine on the M1 chip, analyzing its architecture and discussing its evolution with the M5 chip's integration into GPU cores.
A performance throttle in the Apple M3 Neural Engine was discovered when weight sizes are multiples of 1 MiB, causing throughput drops. By avoiding the problematic DMA path, token throughput for models like Llama 3.2 and Qwen3-8B was significantly improved.
A comprehensive reverse-engineering guide to the Apple Neural Engine, detailing its architecture, programming interface, and performance across A11–A18 and M1–M5 chips, based on direct measurements and static analysis of the private runtime, compiler, and firmware.
A developer reverse-engineered Apple's private APIs to enable training neural networks directly on the Apple Neural Engine (ANE) in M4 Macs and iPhones, bypassing CoreML and GPU. The project demonstrates that ANE hardware is capable of training, though with limitations like low utilization and CPU fallbacks for some operations.
A reverse-engineered implementation enabling neural network training on Apple's Neural Engine via private APIs, demonstrating hardware capability beyond inference.