In 2025, edge AI will officially enter a critical phase of large-scale implementation.According to the latest data from IDC, the edge AI server market in China is expected to grow by 70.5% year-on-year in 2024, with the total market size projected to exceed $1.8 billion in 2025.Gartner predicts that by 2026, approximately 45% of new … Read more
As AI applications migrate to edge devices, AI chips specifically designed for edge computing are becoming the new focal point of competition. This chip revolution is pushing intelligent computing from the cloud to the network edge, ushering in a new era of ubiquitous intelligence. Core Directions of Architectural Innovation Storage-Compute Integration: Breaking the von Neumann … Read more
MCUs (Microcontrollers) and DSPs (Digital Signal Processors) are commonly used hardware components in edge AI devices. This article provides a brief analysis of the characteristics of these two types of devices, in conjunction with the book “AI at the Edge“. A microcontroller is a small, inexpensive computer that is involved in the manufacturing of almost … Read more
GCC NEWS On November 14, 2025, during the “Ascend CANN Open Source Salon – Zhejiang Station” event held in Hangzhou, the Global Computing Coalition (GCC) collaborated with Hangzhou Tiankuan Technology Co., Ltd., Hangzhou Artificial Intelligence Computing Center, and the Zhejiang Ascend Ecological Innovation Center to jointly launch the “CANN Ecosystem Co-Creation Plan,” marking a solid … Read more
Contact me for automotive-grade chip business consultation👆🏻 In the field of smart driving chips,the coexistence of ultra-high computing power (TOPS) and ultra-high functional safety (ASIL-B/D) is one of the core challenges in high-performance SoC design. Automotive-grade AI chips represented by NVIDIA Orin and Horizon Journey 5 achieve the dual goals of performance and safety through … Read more
On September 10, the Arm Unlocked 2025 AI Technology Summit was held. At this summit, Arm officially launched the Arm Lumex Compute Subsystem (CSS) platform, while also announcing the complete lineup of C1 CPU IP and Mali G1 GPU IP. In the afternoon, James McNiven, Vice President of Product Management for Arm’s Client Division, and … Read more
AllwinnerT153Chip Deep Dive: Solidifying the Foundation of Industrial Intelligence with Heterogeneous Computing Power and High Expandability On September 24, 2025, Allwinner Technology made a significant announcement in Shanghai regarding the launch of its next-generation industrial-grade system-on-chip, the T153. As a core product under Allwinner’s “Industrial 2.0” strategy, the T153 is precisely positioned as a “value-for-money … Read more
—————————————————————————————————————————— Copyright Statement: Author: Wang Weiwei. Mainly engaged in ISP/GPU/video/automotive chips/AI chips/SoC architecture design Zhihu Column: Advanced Path of Chip Design WeChat Official Account: Advanced Path of Chip Design (x_chip) Can be forwarded without authorization, but this statement must be retained, otherwise legal action will be taken! —————————————————————————————————————————— Currently, we are in a data-driven intelligent … Read more
—————————————————————————————————————————— Copyright Statement: Author: Wang Weiwei. Mainly engaged in ISP/GPU/video/automotive chips/AI chips/SoC architecture design Zhihu Column: Advanced Path of Chip Design WeChat Official Account: Advanced Path of Chip Design (x_chip) Can be forwarded without authorization, but this statement must be retained, otherwise legal action will be taken! —————————————————————————————————————————— Currently, we are in a data-driven intelligent … Read more
Keywords: on-device generative AI, NPU (Neural Processing Unit), heterogeneous computing, Qualcomm AI Engine, Snapdragon, AI software stack Unlocking on-device generative AI with an NPU and heterogeneous computing https://www.qualcomm.com/content/dam/qcomm-martech/dm-assets/documents/Unlocking-on-device-generative-AI-with-an-NPU-and-heterogeneous-computing.pdf https://www.qualcomm.com/developer/artificial-intelligence This article has 13,299 words and takes 43 minutes to read, podcast 8 minutes Related ArticlesEvaluating the LLM inference performance and energy efficiency of Qualcomm 100 … Read more