AI SoC / Edge AI SoC
Subcategory of integrated SoC designs targeting AI inference, computer vision, and embedded edge computing workloads.
Covers integrated System-on-Chip circuits used as central processing units in embedded systems, robotics, AIoT, and edge AI devices. In this schema, parentCategory is set to 'other' because the parentCategory enum does not include a compute category.
AMD Ryzen AI Embedded P100
AMD Ryzen AI Embedded P100 is an edge SoC with Zen 5/5c CPU (up to 12 cores), RDNA 3.5 GPU, and XDNA 2 NPU — up to 50 TOPS on NPU, 80 TOPS combined, for mobile robotics 2025+.
Apple A11 Bionic
Apple SoC from September 2017 (iPhone 8, 8 Plus, iPhone X), 10 nm TSMC, 4.3 billion transistors, 6 ARMv8.2-A cores. First Apple chip with a dedicated Neural Engine (600 billion ops/s) powering Face ID and Animoji. Discontinued in 2020.
Apple M7 Ultra
Apple-announced server chip planned for H1 2027, supporting up to 1.5 TB of RAM. Descended from the architecture built for the discontinued Project Titan. Apple is skipping M6 Pro/Max/Ultra variants to speed up its debut.
Apple Neural Engine
Apple's series of AI accelerators (NPU), integrated as a block in A- and M-series SoCs since 2017. M4 reaches 38 TOPS; an expanded version is announced in the M7 Ultra server chip for 2027.
D-Robotics Sunrise
D-Robotics' family of proprietary AI-accelerated SoCs for robotics and edge AI; e.g. Sunrise 5 delivers about 10 TOPS in the RDK X5 kit.
Hailo-8L M.2
13 TOPS M.2 AI accelerator drawing only 1.5 W with no external DRAM — designed for Raspberry Pi AI HAT+.
Qualcomm QCS8550
Qualcomm robotics/IoT system-on-chip based on Snapdragon 8 Gen 2. Combines a Kryo CPU, Adreno 740 GPU and Hexagon NPU, delivering up to 48 TOPS (INT8) of on-device AI compute.
XPENG Turing AI Chip
XPENG's in-house AI SoC for advanced AI models in autonomous driving and in-cabin assistants; ~750 TOPS of sparse compute, supporting models up to 30 billion parameters.