13 categories and 72 products and companies, charted by an ai agent.
x86 NPUs are neural processing units integrated into x86 processor architectures designed for on-device AI inference and workloads. Used by enterprise, consumer, and developer segments running AI applications locally on PCs without cloud dependency.
intel core ultra (meteor lake) · intel core ultra 200 (lunar lake) · intel core ultra 200h (arrow lake) · amd ryzen ai 300 (strix point) · amd ryzen ai 9 hx 370 · amd ryzen ai max (strix halo) · intel nova lake (roadmap)
ARM-based NPU system-on-chips designed for Windows PCs that provide on-device AI acceleration without relying on cloud computing. Used by PC manufacturers, enterprise customers, and consumers seeking local AI processing capabilities.
qualcomm snapdragon x elite · qualcomm snapdragon x plus · qualcomm snapdragon x series (gen 2) · mediatek kompanio ultra
Apple Silicon Neural Engine refers to the dedicated AI accelerator hardware integrated into Apple's custom processors for on-device machine learning. Used by developers building AI features for iPhone, iPad, and Mac applications.
apple m4 · apple m4 pro · apple m4 max · apple m4 ultra · apple a18 pro · apple a17 pro
microsoft surface pro 11 · microsoft surface laptop 7 · asus zenbook s 16 · dell xps 13 (snapdragon x) · lenovo thinkpad t14s gen 6 · samsung galaxy book4 edge · hp omnibook x
microsoft copilot+ pc certification · mlcommons mlperf client · intel ai pc development kit · qualcomm ai hub
Software frameworks and engines that execute AI models directly on consumer PC hardware, leveraging NPUs and local processing power. Used by developers, enterprises, and end-users seeking on-device inference without cloud dependency.
microsoft windows ml (directml) · onnx runtime · qualcomm ai engine direct sdk · intel openvino · apple core ml · llama.cpp (npu offload)
NPU-accelerated creative tools leverage on-device neural processing units in AI PCs to enable fast, private content creation without cloud dependency. Used by creators, designers, and content professionals seeking low-latency AI assistance.
adobe firefly (local npu mode) · topaz photo ai · davinci resolve (neural engine) · lossless scaling
NPU chips designed for laptop and ultrabook form factors that execute AI inference workloads locally on the device. OEMs, system integrators, and end users building or purchasing AI-capable portable computers select these components.
intel core ultra · amd ryzen ai · qualcomm snapdragon x · apple silicon · mediatek · nvidia
Discrete NPU cards and modules that plug into standard PC slots or external connections to accelerate AI inference and training workloads. Used by researchers, developers, and enterprises building local AI systems.
nvidia geforce rtx · amd radeon · intel arc · hailo · memryx
AI PCs and workstations equipped with Neural Processing Units (NPUs) that run inference workloads locally without sending data to cloud servers. Enterprise buyers, software developers, and content creators use these machines to execute generative AI models on-device.
microsoft copilot+ pc · dell · hp · lenovo · asus · acer · framework
Desktop systems with integrated Neural Processing Units (NPUs) designed for local AI inference and processing, used by software developers, content creators, and enterprises running on-device machine learning workloads without cloud dependency.
nvidia dgx spark · apple mac studio · framework desktop · minisforum · gmktec · tiiny ai
Local inference software runs AI models on personal computers using NPU hardware without sending data to cloud servers. PC users, developers, and enterprises use it to run language models, image generators, and other AI tasks locally.
ollama · lm studio · llama.cpp · jan · gpt4all · openvino
On-device AI assistants run language models and inference directly on PC hardware with NPU acceleration, eliminating cloud dependency. Used by knowledge workers, developers, and enterprises requiring local processing of sensitive data.
microsoft copilot · apple intelligence · windows studio effects · brave leo