Tagged articles
NPU
34 articles · Page 1 of 1
An engineering teardown of the Arrowspace S1 AI screwdriver reveals how a 2 MP camera, polarized lighting, and a 0.5 TOPS NPU identify screw types in milliseconds, while a 1 kHz strain-gauge sensor and PID loop achieve ±5% torque accuracy — all within a 0.5 W compute budget powered by a single 18650 cell.
Computer VisionEdge InferenceHardware Teardown
0 likes · 13 min read

This article explains why specialized AI chips like Google's TPU and mobile NPUs are needed, details their architectures (systolic arrays, SoC integration), compares generations and vendors (Ascend, Cambricon, Edge TPU, Habana), and covers key metrics (TOPS, TOPS/W) and market landscape across training, inference, and edge.
AI AcceleratorsAI chip marketASIC
0 likes · 11 min read


UbuntuSep 27, 2026 · Industry Insights
Canonical and Qualcomm announce Ubuntu native support for Snapdragon X2 with 18-core Oryon CPU, 80 TOPS NPU, and 15-year maintenance, targeting 2027 release; analysis covers hardware specs, software certification value, ecosystem gaps, and on-device AI implications.
ARM laptopsCanonicalNPU
0 likes · 11 min read

This article explains why CPUs struggle with MoE model prefill but excel at decode, compares dense, SSM, MLA, and BitNet architectures, evaluates CPU, NVIDIA GPU, Apple UMA, and NPU hardware for local LLM inference, and provides a practical tuning guide for running 32B models on a 32GB Mac mini with MLX and Ollama.
Apple SiliconKV CacheLLM inference
0 likes · 21 min read


UbuntuAug 27, 2026 · Artificial Intelligence
The Arduino VENTUNO Q, priced at $299 and shipped with Ubuntu, combines a Qualcomm Dragonwing IQ‑8275 NPU delivering up to 40 TOPS for on‑device large‑language‑model inference with an STM32H5 MCU handling sub‑millisecond real‑time control, offering integrated support for offline voice assistants, multi‑camera vision, ROS 2 robotics, and industrial CAN‑FD interfaces, positioning it as a ready‑to‑use edge AI platform distinct from Raspberry Pi or generic Ubuntu PCs.
ArduinoLLMNPU
0 likes · 11 min read

This guide details HarmonyOS 7's Core Vision Kit ImageSR API for on-device 4x AI super-resolution, covering API usage, ArkTS demo code, permissions, constraints (2048x2048 input, RGBA_8888 format), common pitfalls (emulator unsupported, model download, resource leaks), and suitable use cases like e-commerce and social images where perceptual quality matters more than pixel-perfect accuracy.
API 26ArkTSCore Vision Kit
0 likes · 21 min read

SemiAnalysis’s STEEL lab dissects the Kirin 9030 SoC built on SMIC’s N+3 node, revealing its die layout, CPU/GPU/NPU micro‑architectural changes, performance‑per‑area trade‑offs, and the impact of export‑control‑driven process choices compared with TSMC N6‑based competitors such as the Helio G99, Apple, and Qualcomm.
GPUKirin 9030N+3 process
0 likes · 14 min read

In 2026 China's domestic compute capacity will shift from pilot substitution to large‑scale commercial use, driven by national network projects, East‑West computing phase II and mandatory domesticisation, resulting in three parallel technology routes, layered market segmentation, full‑stack autonomy requirements, booming peripheral components, and a rapid move toward compute‑as‑a‑service.
2026ChinaCompute
0 likes · 5 min read

MoWorld introduces a Flash World Model that runs on domestic Ascend NPU, delivering cinematic‑grade quality at up to 50 FPS real‑time interaction while cutting inference cost by 70%, showcasing a breakthrough in low‑cost, high‑frame‑rate world‑model deployment.
14B MoEFlash World ModelHigh Frame Rate
0 likes · 12 min read

This article breaks down the core chip concepts—CPU, GPU, NPU, APU, SoC, HBM and Chiplet—explaining their functions, key characteristics, historical evolution, and how they relate to each other, and provides a 2026 mainstream‑chip comparison and selection guide.
CPUChipletGPU
0 likes · 18 min read

The article provides a detailed technical analysis of Huawei's Ascend 950 NPU series, covering its one‑chip dual‑structure for training and inference, SIMD/SIMT dual‑mode compute, ultra‑fine memory granularity, PD separation, native FP4 support, a high‑bandwidth 2.0 interconnect, and a fully self‑developed yet CUDA‑compatible ecosystem.
AI acceleratorAscend 950FP4
0 likes · 10 min read

The Tsinghua‑Tencent Mixed‑X team captured the MLSys 2026 MoE inference optimization championship by analyzing NPU bottlenecks, redesigning data movement, applying expert‑level sharding, continuous DMA, PSUM batching, and an Agent‑based optimizer, achieving a 4.1× end‑to‑end speedup while preserving bit‑level output fidelity.
Agent optimizerInference OptimizationMLSys 2026
0 likes · 14 min read

The article breaks down the seven major AI‑focused processors—CPU, GPU, NPU, TPU, LPU, DPU, and VPU—explaining each one's architectural strengths, typical workloads, representative vendors, and trade‑offs, then summarizes which role each chip excels at in modern AI systems.
CPUDPUGPU
0 likes · 9 min read

The AMD Ryzen AI MAX+ PRO 495 (code‑named Gorgon Halo) boosts memory bandwidth, expands unified memory to up to 256 GB, and delivers 55‑60 TOPS NPU performance, resulting in roughly 4 % multi‑core and 3 % single‑core gains over its predecessor while targeting demanding AI workloads on thin‑and‑light laptops.
AMDMobile APUNPU
0 likes · 9 min read

Engineers training large AI models often see noticeable FP16/BF16 result differences between GPUs and NPUs, and even between generations of the same chip, due to floating‑point representation limits, hardware design choices, software library implementations, compiler optimizations, and parallel execution nondeterminism.
AIGPUNPU
0 likes · 10 min read

Huawei’s Da Vinci AI architecture, introduced with the Kirin 810 SoC, combines a 3D Cube matrix‑multiply engine, vector and scalar units, and flexible scaling to deliver high‑performance, energy‑efficient AI compute across devices from low‑power IoT to high‑end cloud servers.
3D CubeAIDa Vinci architecture
0 likes · 11 min read

This article examines Huawei's Ascend AI accelerator architecture, detailing its heterogeneous compute units, memory hierarchy, task scheduling, programming model, and chip variants, while also discussing future challenges and the ecosystem needed for widespread AI deployment.
AI acceleratorAI hardwareDaVinci architecture
0 likes · 14 min read

This article explains what ASIC chips are, how they differ from CPUs, GPUs and FPGAs, classifies them by customization level and function, outlines their performance and cost advantages, discusses their drawbacks, and reviews current products and market trends driving AI hardware adoption.
AI hardwareASICFPGA
0 likes · 11 min read

This article explains the rise of AI‑specific processors, defines AI chips, compares their architectures, and examines the distinct requirements of training versus inference chips while outlining the main technology routes (GPU, FPGA, ASIC) and future outlook.
AI chipsASICDSA
0 likes · 9 min read


JD TechMar 19, 2025 · Artificial Intelligence
This article details JD Retail's AI engine that seamlessly supports both GPU and domestic NPU hardware, describing its heterogeneous cluster architecture, unified training and inference APIs, performance optimizations, extensive model coverage, and multiple production use cases across e‑commerce, logistics, and intelligent assistance.
AI EngineGPUJD Retail
0 likes · 20 min read

JD Retail’s Nine‑Number Algorithm Platform delivers an end‑to‑end AI engine that unifies GPU and domestic NPU resources across a thousand‑card cluster, offering zero‑cost model migration, optimized training and inference pipelines, support for over 40 LLM and multimodal models, and proven business‑level performance that reduces dependence on overseas chips.
AIGPUNPU
0 likes · 19 min read

The Bilibili TTV team optimized OpenSora and CogVideoX text‑to‑video models by redesigning data storage with Alluxio, parallelizing VAE encoding, applying dynamic sequence‑parallel and DeepSpeed‑Ulysses attention, adapting GPU code for NPU execution, leveraging profiling‑driven kernel fusion, FlashAttention, and expandable memory to dramatically increase training efficiency and frame throughput, while outlining future pipeline‑parallel and ZeRO‑3 scaling plans.
FlashAttentionNPUdata pipeline
0 likes · 26 min read

This article describes JD Retail's AI engine built on domestic NPU chips, covering challenges, heterogeneous GPU‑NPU scheduling, high‑performance training and inference engines, extensive model support, real‑world deployment cases, and future plans for large‑scale chip clusters and ecosystem development.
AIGPUNPU
0 likes · 20 min read

This article details JD.com's challenges and solutions for deploying domestic NPU chips across heterogeneous GPU‑NPU clusters, covering architecture, scheduling, high‑performance training and inference engines, real‑world case studies, and future plans to scale AI workloads securely and efficiently.
AIJD.comNPU
0 likes · 19 min read

This guide walks through the fundamentals of Huawei Ascend NPU hardware, the CANN software stack, driver and firmware installation, Kubernetes integration via Docker runtime and device plugin, and a complete ResNet‑50 inference demo on Ascend 310P.
AI inferenceCANNDocker Runtime
0 likes · 12 min read

The article examines the rapid growth and increasing complexity of edge AI NPUs, discussing challenges in software and hardware acceleration, supply‑chain constraints, and the need for integrated engine solutions to sustain performance and power efficiency.
NPUedge AIsoftware complexity
0 likes · 9 min read

The article explains what Neural Processing Units (NPUs) are, how they differ from CPUs and GPUs, their parallel architecture, the workloads they accelerate, their role in edge AI and AI‑enabled PCs, and why industry analysts expect NPU‑enabled devices to dominate the market by 2026.
AI PCAI acceleratorEdge Computing
0 likes · 8 min read

The report analyzes the accelerating shift from x86 to ARM in AI‑enabled devices, covering architectural differences, market share dynamics, Apple’s successful ARM transition, Microsoft’s ARM ecosystem, Intel’s heterogeneous AI processors, rising memory demands, and future industry forecasts for 2024‑2027.
AI PCARMNPU
0 likes · 17 min read


21CTOMay 29, 2024 · Artificial Intelligence
Microsoft’s vision of AI PCs, highlighted by the Copilot+ concept, details how integrated NPU hardware, local large‑language models, and the Windows Copilot Runtime enable on‑device AI inference, reducing data‑center load and offering developers a unified platform for building next‑generation AI applications.
AI PCCopilot+Edge Computing
0 likes · 11 min read

The article examines how CPUs, GPUs, and memory have long formed the core of PC hardware, discusses the emerging role of NPUs for AI processing, and describes Intel's Lunar Lake strategy of integrating memory with the processor to deliver faster, lower‑latency performance in upcoming AI‑focused PCs.
AI PCCPUGPU
0 likes · 5 min read

The article explains the various AI‑capable chips such as CPUs, GPUs, FPGAs, NPUs, and TPUs, compares their performance and efficiency, describes heterogeneous CPU+xPU solutions, and provides market share data while highlighting the growing adoption of specialized AI accelerators.
AI accelerationAI chipsCPU
0 likes · 7 min read

The paper presents a holistic algorithm‑system‑hardware co‑design for advertising deep‑learning inference, combining model pruning, approximate computing, kernel fusion, scheduling and PCIe transfer optimizations with GPU and NPU upgrades, achieving up to five‑fold speed‑up and significantly higher latency‑bounded QPS for large‑scale ad services.
Algorithmic OptimizationGPUNPU
0 likes · 24 min read

Edge deep learning inference on mobile devices faces hardware and software fragmentation, diverse CPUs, GPUs, DSPs, and NPUs, and limited programmability; optimization techniques such as model selection, quantization, and architecture‑specific tuning enable real‑time performance, with most inference on CPUs, GPUs offering 5–10× speedups, and co‑processor support varying across Android and iOS.
DSPEdge InferenceGPU programming
0 likes · 17 min read

This article explains heterogeneous computing and compares major processing units—CPU, GPU, FPGA, ASIC, and NPU—highlighting their architectures, strengths, and typical use cases, especially in deep‑learning and AI workloads.
ASICCPUFPGA
0 likes · 10 min read
