Huawei, the Chinese tech giant, has revealed an ambitious strategy to advance AI-powered solutions across diverse industries and strengthen its computing infrastructure to cater to global individual, corporate, and government consumers.
In a keynote address, Sabrina Meng, Huawei’s Deputy Chairwoman, CFO, and Rotating Chairwoman, unveiled a whitepaper titled “Accelerating Intelligent Transformation.” Meng emphasized the company’s commitment to delve into foundational AI technologies and create a robust computing foundation for China.
The company aims to enhance the synergy between hardware, software, chips, edge computing, devices, and cloud services to foster a thriving ecosystem.
Huawei’s ultimate objective is to address the diverse AI computing requirements of various industries, positioning itself as a global leader.
As artificial intelligence gains prominence worldwide, Huawei’s “All Intelligence” strategy is designed to assist industries in seizing strategic opportunities presented by AI.
At the Huawei Connect event, the company released a reference architecture for facilitating intelligent transformation, offering practical guidance and references to help industries leverage intelligence effectively.
About Huawei Connect
Huawei Connect is an annual flagship event in the global ICT industry where Huawei unveils innovative solutions.
Over the years, Huawei has used this platform to envision an intelligent world, set goals to become a major cloud service provider, and unveil strategies related to AI and computing.
Due to the COVID-19 pandemic, this year’s event marked the return to physical presence in Shanghai after a virtual edition in 2020.
The event brought together leaders, experts, partners, developers, and industry stakeholders from around the world to explore opportunities in the intelligent future.
In a related keynote session, David Wang, Huawei’s Executive Director of the Board and Chairman of the ICT Infrastructure Managing Board, introduced the Atlas 900 SuperCluster, the latest addition to Huawei’s Ascend series of computing products.
This supercluster is optimized for training massive AI models with over one trillion parameters, featuring high-density 800GE ports and enhanced system reliability. These improvements allow for more efficient continuous model training.
Wang also highlighted Huawei’s strengths in computing, storage, network, and energy, emphasizing the importance of system reliability for training massive foundation models.
Pangu AI Model
Huawei introduced the Pangu AI model, which addresses specific customer needs across various sectors, such as mining, government, vehicles, weather, medicine, virtual humans, and R&D.
This extension of the Pangu model enhances AI model development and significantly shortens training cycles for complex scenarios, like autonomous driving.
The Pangu model also includes CodeArts Snap, an intelligent programming assistant for developers, capable of generating code, suggesting code, commenting, generating test cases, and deploying services and apps efficiently.
The company launched the Huawei Cloud Ascend AI cloud service, available for purchase on the Huawei Cloud website. It also introduced a dedicated zone for foundation models of Ascend AI cloud service, offering open-source models like LLaMA and GLM. The service supports popular AI frameworks, simplifying migration for operators.
Huawei Cloud CTO, Bruno Zhang, announced a long-term research project to innovate software and hardware structures to better support AI deployment, making Huawei Cloud the preferred choice for the AI era.
The focus is on enhancing system architecture, data processing, model training, application development, and AI cloud service layers. Ascend AI cloud service is now accessible for mainstream use, enabling enterprises to deploy versatile AI platforms in existing data centers.
Huawei’s comprehensive approach to AI and computing demonstrates its commitment to becoming a global leader in these fields and providing innovative solutions across diverse industries.