Computing power has become a new productive force, driving social and economic development, and the explosive growth in demand for computing power has driven the computing industry into an era of diverse computing. As the computing power landscape accelerates transformation, ARM's computing power advantage extends to the server market. Facing the era of diverse computing, Huawei joins hands with industry partners to jointly build a computing industry ecosystem. Through four major strategies: hardware openness, software open source, enabling partners, and developing talents, Huawei achieves openness of full-stack capabilities and enables innovation among partners across the industry. Core technology research and development + motherboard openness create an open Kunpeng computing hardware industry, put partners first, and enable partners to innovate with differentiation. Lead open source + contribute to open source, and prosper the basic software ecosystem. With all-round partner enablement and industry-wide innovation power gathered, there are already 800,000 developers, 3,500+ partners, and 10,000+ solutions certified by Kunpeng. Carry out in-depth talent training in colleges and universities to provide a steady stream of talent power for industrial development.

Based on the HUAWEI Ascend series of AI processors and basic software, the Atlas artificial intelligence computing solution is built, including Atlas series modules, boards, small stations, servers, clusters and other rich product forms, to create a solution for "end, edge and cloud". "Full-scenario AI infrastructure solution covering the entire process of reasoning and training in the field of deep learning.

Kunpeng general computing platform provides TaiShan servers, Kunpeng motherboards and development kits based on Kunpeng processors. Hardware manufacturers can develop their own brand products and solutions based on Kunpeng motherboards; software manufacturers can develop application software and services based on openEuler open source OS and supporting database, middleware and other platform software; Kunpeng development kits can help developers accelerate application migration and computing Power upgrade. Kunpeng general computing platform adapts to the diverse computing and green computing needs of various industries and is committed to building the strongest computing power platform.

Atlas 900 PoD (Model: 9000) is the basic unit of the AI ​​training cluster based on Huawei Ascend and Kunpeng processors. It has powerful AI computing, best AI energy efficiency, and best AI scalability. Cluster basic units are widely used in deep learning model development and training scenarios, and are ideal for computing-intensive industries such as smart cities, smart medical care, astronomical exploration, and oil exploration.

The Atlas 800 inference server (model: 3000) is equipped with an Ascend processor and supports up to 8 Atlas 300I inference cards, providing powerful real-time inference capabilities. It is widely used for AI inference in data centers.

Atlas 200 DK (Model: 3000) is a high-performance AI application development board that integrates the Ascend processor to facilitate rapid development and verification. It has been widely used in scenarios such as developer solution verification, higher education, and scientific research.

Atlas 300I inference card (model: 3000/3010) is equipped with Ascend processor, enabling excellent AI inference performance. A single card has strong computing power and supports 80 channels of real-time high-definition video analysis. It is an ideal choice for intelligent scenarios such as smart cities, transportation, and finance.

Atlas 500 AI edge station (Model: 3000) is designed for edge applications. It has the characteristics of small size, strong environmental adaptability, easy maintenance, cloud-edge collaboration, etc., and can be widely deployed at the edge. Atlas 500 AI edge station meets the complex needs of security, transportation, communities, campuses, shopping malls, shopping malls, supermarkets and other scenarios.