Artificial intelligence (AI) is reshaping technology at an unprecedented speed and becoming an indispensable part of people's daily lives. The Arm computing platform is at the core of this transformation. The cumulative shipment of chips based on Arm architecture has exceeded 310 billion, widely used in consumer electronics devices, AI cars, and AI first data centers. On the eve of last week's Taipei COMPUTEX exhibition, Chris Bergey, Senior Vice President and General Manager of the Terminal Business Unit of Arm, delivered a brilliant presentation on the rapidly developing AI revolution in Arm's keynote speech and had in-depth conversations with senior executives from two key partners - MediaTek and NVIDIA. As Bergey said, AI is changing everything at an unprecedented speed. Arm and its world-class ecosystem continue to innovate and invest in hardware and software to achieve the next wave of AI that will change the lives of billions of people. Arm computing platform becomes the innovative core of Taipei COMPUTEX 2025 Over the years, Arm's product portfolio has continued to evolve to meet the needs of its target market. Nowadays, Arm has transformed into a computing platform company, focusing on helping partners integrate technology more quickly and scale up through a system level approach to seize the opportunities brought by AI. From cloud to edge applications, Arm is demonstrating leadership in computing energy efficiency and energy efficiency. In terms of cloud and data centers, as AWS, Google, and Microsoft continue to expand their self-developed data center chips equipped with the Arm computing platform, Arm expects that by 2025, nearly 50% of the computing power shipped to top ultra large scale cloud service providers will be based on the Arm architecture. This growth momentum comes from the strong demand in the industry for Arm's efficient and energy-saving computing capabilities to support complex AI inference and training workloads. Thanks to Arm's insistence on "per watt performance", the top cloud service providers can achieve up to 40% energy efficiency improvement using Arm architecture chips compared to other platforms. At this year's exhibition, NVIDIA also shared the latest development momentum of NVIDIA Grace CPU based on Arm architecture, which has provided performance and efficiency improvements for demanding AI workloads in a series of practical deployments including ExxonMobil, Meta, etc. Performance per watt is also crucial in edge devices. At last year's Taipei COMPUTEX exhibition, Arm released the Arm Terminal Computing Subsystem (CSS) specifically designed for consumer electronics devices, serving as the computing platform for flagship AI smartphones and next-generation AI PCs. In addition to bringing double-digit performance improvements, terminal CSS also brings substantial benefits to users - faster application launch, smoother and more durable AI experience. Currently, 99% of smartphones worldwide use the Arm architecture, and the demand for performance, battery life, and "always on" is constantly expanding to other consumer device markets. In fact, the design of AI PCs is increasingly approaching the form of modern smartphones - with slim and light dimensions, fanless design, all-weather battery life, always online efficiency, and excellent multimedia experience for both video conferencing and video streaming. In recent years, the demand for Arm architecture in the PC and tablet markets has grown significantly, and it is estimated to account for over 40% of overall shipments by 2025. At present, most mainstream applications worldwide have launched Arm native versions for Windows, bringing faster and more powerful AI PC experiences, covering core applications such as chatbots and productivity tools. MediaTek's recently launched Kompanio Ultra SoC further enhances the level of the Chromebook market, enabling the new generation of Chromebook Plus devices to have advanced AI and multimedia capabilities. In addition, NVIDIA's DGX Spark AI desktop computing platform is also showing strong momentum. The platform is equipped with Grace Blackwell supercomputer based on Armv9 architecture CPU, which has AI computing power supporting 200 billion parameter models, and has attracted the attention of developers and researchers who are committed to building powerful next-generation AI models. At this year's Taipei COMPUTEX exhibition, NVIDIA announced its partnership with leading global system manufacturers such as Acer, Asus, Dell Technologies, Gigabyte, HP, Lenovo, and MSI to create NVIDIA DGX Spark and DGX Station systems. Breakthrough the computing limit with Armv9 architecture Arm's ability to penetrate these diverse markets is thanks to its Armv9 architecture, which has become the foundation of modern computing. Since its launch four years ago, many of the latest smartphones and PCs worldwide have adopted Armv9 CPUs, bringing even better AI performance to these devices. Arm is expected to release a new Armv9 flagship CPU (codenamed Travis) later this year. It will bring double-digit performance improvements and further accelerate the processing of AI workloads through the latest architecture feature - Scalable Matrix Extension (SME). At the same time, Arm's next-generation GPU (codenamed Drage) will be able to provide sustained performance support for long-term gaming and richer multimedia content. The combination of these two products will introduce the Arm Lumex CSS solution for the mobile market, laying the foundation for edge AI performance on future consumer electronics devices. Accelerating the development of Arm architecture AI workloads from the cloud to the edge In addition to innovative hardware, Bergey also reviewed Arm's globally leading software ecosystem. Currently, over 22 million developers are developing on the Arm platform. It is precisely because the development community is so strong and diverse that Arm continues to invest heavily in software to assist developers in reducing complexity and improving efficiency during the development process. At last year's Taipei COMPUTEX exhibition, Arm released the Arm Kleidi software library, allowing developers to achieve real-time optimal performance on various AI models and workloads, whether they are audio, images, text, or video. Within just one year of Kleidi's launch, it has integrated with multiple mainstream AI frameworks, including ExecutuTorch, PyTorch, Angel, llama.cpp, MediaPipe, MNN, and ONNX Runtime. So far, Kleidi has accumulated over 8 billion installations on devices equipped with Arm architecture and is still growing. The ongoing AI transformation is a once-in-a-lifetime opportunity Looking ahead to the future, the pace of AI's continuous evolution and innovation will never stop. In the past 18 months alone, over 150 basic models covering speech, image, text, and video generation have been developed. AI assistants that originally only existed in the cloud are now being prioritized for development on edge devices, driving AI inference tasks from the cloud to edge applications. Despite significant investment in training, inference (the actual process of AI operation and value creation) is the key driver of future innovation and business value. AI agents are also growing rapidly, as these systems are capable of independently executing complex tasks, collaborating with each other, and automatically running on a large scale. This will become the foundation of physical AI, extending AI from the digital world to real physical environments such as robots. As we can see today, scalable and efficient computing will become even more important in the future. Bergey emphasized in his keynote speech at Taipei COMPUTEX that the AI era is a once-in-a-lifetime opportunity to redefine how technology impacts the world. But to truly realize this potential, it still relies on continuous innovation and deep collaboration with Arm's world-class ecosystem. From the cloud to the edge, Arm's computing platform is at the core of this transformation, not only driving AI everywhere, but also shaping its future step by step. And Arm is the cornerstone of the future. This is reported by Top Components, a leading supplier of electronic components in the semiconductor industry They are committed to providing customers around the world with the most necessary, outdated, licensed, and hard-to-find parts. Media Relations Name: John Chen Email: salesdept@topcomponents.ru