How Huawei is Competing Against Nvidia with GenAI

The global battle for dominance in generative AI is heating up, with technology giants vying for dominance in this transformative field. U.S. companies like OpenAI, Nvidia, and Google are leading the way, while Chinese competitors like Baidu and Alibaba are on the rise, with another major player, Huawei, also making great strides. Recent developments and strategy announcements from the Chinese tech giant point to a strong and comprehensive approach to GenAI that spans all aspects from hardware to software to cloud services.

Global executives should take note of Huawei’s strategic approach to AI, which offers valuable lessons for staying competitive.

Strategic integration across the stack: Huawei’s holistic approach spanning hardware, software, and cloud provides a blueprint for establishing leadership in emerging technologies. Ecosystem collaboration is another key element, with partnerships enabling us to accelerate AI adoption and innovation and provide tailored solutions for specific industries. In addition, staying ahead of geopolitical and regulatory challenges is also essential, as these can have a significant impact on AI deployment and market leadership. Finally, embedding AI deeply into core business processes is key to future-proofing operations, enabling enterprises to anticipate and shape market trends.

These strategies will not only enhance your current operations, but also position your company as a leader in the evolving AI environment.

Muscle: Increasing AI Computing Power

At the core of Huawei’s GenAI ambitions is its Ascend AI computing infrastructure, which demonstrates its capabilities by training large-scale language models with hundreds of billions of parameters. Huawei’s latest development, the Ascend 910C chip, has already attracted significant interest from Chinese tech giants, with pre-orders exceeding $2 billion. The company claims that the training efficiency of its AI chip not only matches but exceeds that of Nvidia, and performance metrics have already surpassed Nvidia’s A100 chip. Nearly 50% of China’s large-scale language models employ Ascend technology. This computing power establishes a critical foundation for Huawei’s GenAI efforts, propelling the company to greater heights.

Brain: Pangu model 5.0

Huawei’s Pangu Model 5.0 excels at processing tasks across text, images, videos, and voice, and adapts to different languages ​​and scenarios. Pangu’s applications span a wide range of fields, including autonomous driving, industrial design, meteorology, and medicine. Its ability to revolutionize industrial design and significantly shorten the design process of new cars demonstrates its potential for transformation across industries. In particular, the Pangu weather model is the first AI model to surpass traditional numerical forecasting methods in accuracy, providing global weather forecasts within seconds.

Soul: HarmonyOS NEXT

HarmonyOS NEXT represents a pivotal evolution in Huawei’s efforts for an independent, AI-enhanced mobile operating system. By integrating the computing power of Ascend with the capabilities of the Pangu Model, HarmonyOS NEXT features AI-driven features such as AI drawing, AI removal and AI voice repair. These innovations not only enhance the user experience, but also push the boundaries of accessibility and interaction, making the operating system a key element of Huawei’s GenAI strategy.

Cloud: Huawei Cloud Services

Huawei Cloud plays a key role in supporting Huawei’s GenAI ecosystem. To address major challenges in training large-scale language models, such as the “computing wall”, “memory wall” and “energy efficiency wall”, Huawei has launched Elastic Memory Storage Service, which leverages patented memory pooling technology to overcome these barriers. With the establishment of major AI computing centers and an extensive AI computing resource pool, Huawei Cloud is strengthening its role as the backbone of GenAI development.

Ecosystem Collaboration: Building the Future Together

Huawei is building the GenAI ecosystem through strategic partnerships with industry leaders such as Mianbi Intelligence, Honor, and ByteDance. These collaborations are designed to explore vertical model applications across different domains, from intelligent mobile office solutions to large-scale model innovation through terminal-cloud collaboration. The launch of the Huatu Chinese medicine model and Pangu mining model, jointly developed with Shandong Energy Group, demonstrates the industry-specific applications of Huawei’s GenAI technology.

Staying ahead with GenAI

Huawei’s GenAI strategy embodies the company’s ambition to lead this transformational field. From foundational AI computing capabilities, as exemplified by the Ascend 910C chip, to innovative applications, its comprehensive approach has positioned the company as a global leader. For senior executives, lessons from Huawei’s approach highlight the critical importance of full-stack innovation, strategic collaboration, and long-term vision to secure a leadership position in the rapidly evolving AI field.