⬅Back to AI Roadmap
📱
PARALLEL SPECIAL TRACK
⚡ Edge / On-Device AI Parallel Track
Spanning Layer 01 to 11: mobile NPU hardware, Core ML / llama.cpp runtimes, small language models (SLM), and autonomous edge devices.
1
Hardware & Edge NPU
Apple Silicon (Neural Engine)Snapdragon X EliteIntel Core Ultra (NPU)Edge GPU (Jetson Orin)Robotics IoT Chips
2
Runtime & Compilers
llama.cpp (GGUF)Core MLONNX RuntimeTensorRT-EdgeTFLiteWebGPU / WebLLMExecuTorch
3
SLM & Mobile Models
Phi-3.5 / Phi-4 MiniQwen2.5-0.5B/1.5B/3BLlama-3.2-1B/3BSmolLM2Small VLM (PaliGemma)
4
Edge Applications
Smartphone AI (Apple Intelligence)AI PC Copilot+Smart WearablesAutonomous DronesIndustrial IoT Vision