On November 13, 2025, Arm Technology (China) Co., Ltd. officially launched the next-generation NPU IP – “Zhouyi” X3 in Shanghai. Image | Highlights of the “Zhouyi” X3 NPU IP; Source: Arm Technology As the first major product under the “All in AI” product strategy, the “Zhouyi” X3 is based on the latest DSP+DSA architecture designed … Read more
Introduction This article has been in the works since September, and I finally clarified my thoughts this month, presenting some trade-offs. It has gone through three versions, and I still feel something is lacking; I hope for your guidance. I have been pondering a question: what constitutes a good architecture? Our first-generation architecture was borrowed … Read more
Click the blue text to follow us Translation by YoToo In the current wave of technological advancement, 3D printing technology shines like a brilliant new star, igniting an innovative revolution across various fields with its unique charm and limitless potential. When 3D printing cleverly integrates with other domains, it sparks astonishing innovations that bring unprecedented … Read more
What are the differences between CPU, GPU, and NPU, which differ by just one letter? Why hasn’t the CPU been eliminated? Why can’t GPU and NPU replace each other? Today, the instructor will break it down from the “architecture level” so that you can understand it without needing to know chip principles. 1. Role Division: … Read more
Perched on the cliffs of the eastern coast of Mallorca, Spain, lies a seemingly simple yet exquisitely crafted residence—Can Lis.This is a hidden sanctuary created by architect Jørn Utzon for himself and his family, yet it has been acclaimed by numerous architectural critics and peers as one of the most beautiful houses of the 20th … Read more
The core of the NPU compilation process is the “operator”. An operator can be understood as each computational step in the model, such as convolution, pooling, and fully connected layers. The compiler determines the destination of each operator through a multi-stage decision process.The core process of separating tasks between the CPU and NPU can be … Read more
Groq’s Stunning Emergence Time flies, and many of you may remember that last year, in January and February, an AI company with a rather peculiar name, Groq, made headlines with its self-developed LPU (Language Processing Unit) chip system, showcasing extreme low-latency and high-performance inference on the Llama2 70B model (with actual performance exceeding GPU clusters … Read more
The Linux memory management subsystem is a complex and powerful system, whose main goal is to efficiently manage system memory resources, providing stable and reliable memory allocation and reclamation mechanisms for user-space programs and the kernel itself, while also taking into account the characteristics of different hardware platforms. 1. Overall Architecture of the Memory Management … Read more
For nearly fifty years, Intel’sx86 architecture has dominated the computing field, providing core computing power for various devices ranging from entry-level laptops to top-tier data center servers. However, a quiet revolution has been brewing behind the scenes: the Arm architecture, which originated in the smartphone era—specifically, the ARM7 TDMI that powered the Nokia 6110 in … Read more
