Cisco wants to solve the ‘bottleneck’ of AI infrastructure in Vietnam

“Vietnam is facing an important turning point. When businesses transition from testing to applying AI in practice, security and governance elements become key factors,” Associate Professor. Dr. Hoang Huu Hanh, Deputy Director of National Digital Transformation – Ministry of Science and Technology, said at the Cisco Connect Vietnam event on May 5 in Hanoi. “We are building the national infrastructure needed to realize and expand this vision, and look forward to working with partners to make Vietnam a regional AI hub within the next four years.”

At the event, speakers said AI is forecast to contribute an additional 130 billion USD to Vietnam’s GDP by 2040. However, to realize its potential, Vietnam needs to overcome infrastructure “bottlenecks” such as fragmented data, lack of uniform management, as well as gaps in human resources.

“The AI ​​readiness journey starts with network infrastructure,” shared Mr. Nguyen Nhu Dung, General Director of Cisco Vietnam, Cambodia and Laos. “Our vision is to provide a core, secure infrastructure platform for Vietnam to maximize its potential and help businesses confidently enter the AI ​​journey.”

According to Cisco, currently 62% of businesses in Vietnam face the risk of losing value from AI investments due to “AI infrastructure debt” – patchy solutions, lack of uniformity or technical compromises when deploying AI but without a solid foundation. This puts them at risk of not being able to extract value from AI due to the burden of technical debt.

Cisco’s strategic investments in infrastructure, human resources and digital skills aim to accelerate AI adoption and unlock value for businesses. In particular, Cisco Unified Edge, an AI-ready security network architecture, aims to unify connectivity, security and management capabilities in a single platform.

Specifically, Cisco Unified Edge allows businesses to build network infrastructure that can handle large volumes of data from AI applications, while increasing protection against increasingly complex cybersecurity threats. The model currently integrates a series of technologies such as SD-WAN, Zero Trust, AI-Driven Security and centralized management on the cloud platform, helping businesses expand their systems flexibly, optimize access performance and reduce complexity in network operations.

In Vietnam, Cisco also provides Silicon One G300 – a new line of network processor chips under the Silicon One architecture launched in February, designed to serve gigawatt-scale AI data centers and large network infrastructure. The chip aims to handle extremely high data traffic with bandwidth up to tens of terabits per second, reducing power consumption and transmission latency.

In addition, through expanding cooperation with Nvidia in the Secure AI Factory model and implementing the Zero Trust model for AI agents, Cisco said Vietnamese businesses can deploy secure and highly resilient AI.

The company also operates the 5G Routed Optical Network Laboratory (5G RON) in Hanoi, which is currently an important testing environment for carriers and businesses. Partners are now starting to use this lab to test compatibility, optimize AI application scenarios and accelerate network transformation.

Vietnam is currently attracting many global “giants” to invest in AI and semiconductor infrastructure. On May 4, Google Labs and Saigon AI Hub announced they became strategic partners, aiming to deploy Google Labs’ first Applied AI Lab in Vietnam.

Meanwhile, on May 5, Intel completed handing over 31 chip assembly and testing equipment to Ho Chi Minh City High-Tech Park (SHTP) and Hanoi National University for semiconductor training and research purposes. The move takes place in the context that Vietnam aims to develop the semiconductor industry, gradually mastering technology and training tens of thousands of highly qualified engineers. The introduction of real equipment into teaching and research is expected to help shorten the gap between training and production, especially in packaging and testing.