Huawei Yang Chaobin: Industry Intelligence Acceleration Entering Fast Lane, Joining Partners to Cultivate Three Major AI Core Scenarios

Sina Tech News, March 22 — At Huawei China’s Partner Conference 2026, Huawei’s Executive Director and ICT BG CEO Yang Chao-bin stated in his keynote speech titled “Joining Hands with Partners to Build a New Era of AI and Intelligent Infrastructure” that Huawei has collaborated with partners to develop and implement multiple industry-specific scenario solutions. Combining these efforts with exploration in AI application practices, Huawei is working with partners to deeply cultivate three core scenarios: rapid lightweight AI applications, AI reshaping core businesses, and AI-driven innovation models, breaking through industry intelligent transformation.

In the rapid lightweight AI application scenario, Huawei has fully upgraded the Ascend all-in-one machine, with tiered and classified configurations to meet different model needs. Additionally, Huawei has launched a full-stack open-source reference design from hardware to application enablement, lowering the barriers to AI application development. Partners can focus on business integration and knowledge base construction to quickly create scenario-based applications. Currently, applications from over 350 partners have been deployed in key sectors such as education, healthcare, finance, and government-related fields.

For AI reshaping core business scenarios, to meet the key demands of AI Agents operating deeply within business cores—such as low latency and high throughput—Huawei offers three pathways: rapid deployment, compatibility and adaptation, and autonomous development. Leveraging Ascend supernodes and open-source large EP, Huawei helps partners independently build low-latency, high-performance inference services.

In the AI-driven innovation model scenario, Huawei has built a complete reinforcement learning post-training process based on Ascend hardware and software. This includes comprehensive support from base models, RL frameworks, to toolchains, along with one-click deployment RL Docker images and end-to-end RL deployment guidance. Partners can efficiently complete each post-training stage, deeply explore core industry scenarios, and help customers develop industry-specific models to enable intelligent industry transformation.

Meanwhile, Huawei has collaborated with partners to create a 10-gigabit campus network foundation integrating Wi-Fi+X sensing, with comprehensive upgrades in connectivity, sensing, security, and operations. Through southbound container openness and northbound API interfaces, Huawei supports partners in developing scenario-based AI applications such as smart security, smart buildings, and smart healthcare, accelerating AI deployment and value realization.

Yang Chao-bin emphasized, “Industry intelligence is entering an acceleration phase, which is not only a technological upgrade but also a reshaping of business logic. Huawei will focus on core ICT technologies, strengthen the Ascend and Kunpeng computing infrastructure and 10-gigabit campus network foundation, and enable more innovation through open source. With accumulated technology, we empower more industry applications. We aim to make models truly operational in business, creating value for every customer. In the future, Huawei will continue working with partners to develop more scenario-based solutions and seize opportunities in industry intelligence.”

At the conference, Huawei Cloud CEO Zhou Yuefeng delivered a keynote titled “Walking with Huawei Cloud to Seize Opportunities in the AI Era.” He pointed out that Huawei Cloud will center around industry intelligent agents, providing AI infrastructure, large models, and industry intelligent agent development platforms for partners, developers, and enterprise customers to jointly create an “Industry AI Dream Factory” and build a fertile ground for enterprise-level AI innovation.

Huawei Cloud is reshaping AI infrastructure based on intelligent agents, creating an efficient, secure, and reliable AI computing foundation. Committed to self-developed models, open-source SOTA models, and enhanced post-training, Huawei Cloud currently supports over 160 top industry models for immediate use. Additionally, Huawei Cloud has built a rich industry intelligent agent development platform. Since the public beta of CodeArts in February, user numbers have increased sevenfold, and daily active users have grown twelve times. The enterprise version of AgentArts, Huawei Cloud’s one-stop enterprise intelligent agent development platform, will begin public testing in April, with the open-source version of openJiuwen scheduled for May. During this conference, Huawei Cloud also announced a comprehensive upgrade to partner policies and launched a 200 million yuan AI ecosystem fund. Zhou Yuefeng stated that Huawei Cloud will increase support across funding, technology, policies, and platforms to help partners sustain growth.