Project trains neural networks on Apple Neural Engine via reverse-engineered APIs
maderix · github · 2026-07-29
A project trains neural networks on Apple Neural Engine using reverse-engineered private APIs
maderix/ANE is a project for training neural networks on Apple Neural Engine by using reverse-engineered private APIs.
- Language: Objective-C
- Current scale: 7.1k stars and 955 forks
- The project suggests practical experimentation with Apple’s on-device AI hardware rather than a standard consumer app
More from Embodied
- Humanoids are an idealized image, but not an optimal form — wavefnx · 2026-07-29
- DeltaIntelligence raises nearly RMB 500 million to build humanoid foundation models — 快鲤鱼 · 2026-07-29
- ROS 2 TensorRT node turns Depth Anything V3 into real-time depth and point clouds — tom_doerr · 2026-07-29
- NVIDIA Isaac Sim can load 3DGS scans, but a single prim tops out at 16.7M points — rsasaki0109 · 2026-07-29
- Liquid AI plans a talk on its upcoming edge agentic model and harness design — maximelabonne · 2026-07-29
- Four Major Bottlenecks for AI Hardware Companies: Marketing Data Silos and Shifting Needs — 创业邦 · 2026-07-29