Vietnam is shifting to AI

Vietnam's digital economy is expected to reach US$45 billion by 2025, according to Google and Temasek's e-Conomy SEA 2023 report, with AI technology playing a major role in this growth.

Robots that make coffee according to customers' requests demonstrate at a technology event in Vietnam - Photo: DUC THIEN

That was the affirmation of Mr. Phung Viet Thang, Vietnam country director, business, marketing and communications division of Intel Corporation, when sharing with Tuoi Tre about how businesses apply artificial intelligence (AI) as well as the role of AI in the development of Vietnam's digital economy.

Mr. Thang said: With the incredible development speed of AI technology in recent years, we believe that the world is entering a new industrial revolution, an era where intelligent machines with the ability to detect diseases or create art can be seen everywhere.

* How will AI contribute to the growth of Vietnam's digital economy?

Mr. Phung Viet Thang

- Similar to previous industrial revolutions driven by specific elements such as steam, water and electricity, the rise of the AI ​​era will be led by computing power in data storage and processing, which plays an essential role in keeping global information systems from being paralyzed and supply chains from being disrupted...

Widespread deployment of AI depends entirely on the availability and efficiency of computing power.

Therefore, a diverse computing infrastructure plays a crucial role, not just when it comes to data center computing.

A country cannot rely solely on a single dam for water storage and distribution. Likewise, the AI ​​era requires a diversity of computing power from an ecosystem that includes data centers, PCs, and edge computing devices.

* Is the data center the most important role in promoting AI in Vietnam?

- There is no denying the fact that data centers are at the heart of the AI ​​revolution. Gen AI applications, built on large language models (LLMs), require huge amounts of data for training.

This data-intensive processing requires a large-scale, high-performance computing infrastructure with hundreds or thousands of central processing units (CPUs), graphics processing units (GPUs), accelerators, and network processors.

In Vietnam, the data center market value is forecast to increase to US$1.04 billion by 2028, from US$561 million in 2022, achieving a compound annual growth rate (CAGR) of 10.7% according to Savills' update in May 2024.

However, to push the AI ​​revolution further, we need to consider three important factors: economics, geographic distance, and current regulations.

* Could you be more specific about the role of the three factors: economic, geographical distance and current regulations?

- Economically, the cost of running AI-related activities through data centers or cloud computing is quite expensive.

Whether owning or leasing data centers or using cloud computing services, organizations face financial, operational and management challenges.

Furthermore, the geographical distance between where the data is generated and the data centers can cause latency, affecting the performance of applications that require real-time processing, such as self-driving cars.

Finally, not every organization or country is comfortable with their data being stored outside of on-premises servers or national territory due to security and information privacy concerns.

As in Vietnam, Decree 53 of 2022 of the Government requires Vietnamese enterprises and foreign enterprises with branches and representative offices in Vietnam to store data locally on domestic servers.

* More and more businesses and organizations in Vietnam are applying AI to their operations. What are the criteria for choosing the right AI solution, sir?

- AI is a complex technology. Depending on the use case, businesses will need many different computing requirements, including user experience, operational considerations, costs, legal regulations, etc.

To develop sustainable AI, it is necessary to research the most suitable infrastructures to meet the increasing demand for computing power.

Through studying the models, it can be seen that there is no one model that can be applied to all cases, especially in the computing field of the AI ​​era.

In other words, we need a diverse infrastructure to provide computing power that matches demand, because there is no one-size-fits-all solution.

Mr. Vu Thanh Tung (Director of Business and Product Development, GreenNode, VNG):

Businesses must determine their needs and goals for using AI.

Mr. Vu Thanh Tung

The first thing is that businesses need to clearly define what goals they want to achieve and what they need, then determine which problems to prioritize in their business operations.

Some businesses will want to choose difficult problems that can bring breakthrough results, but there are also businesses that prioritize solving small and safe problems first.

This is up to the "taste" of business leaders.

There are many options and flexible solutions on the market, with fairly accessible entry levels.

However, a very important condition for applying AI is whether the enterprise's database meets the requirements or not. Meeting the requirements means where the data is located, whether there is a lot of junk data/valueless data mixed in or not.

Does the data clearly reflect the business processes of the enterprise? And after the data is analyzed, does the enterprise have enough tools to input market feedback back into the data management system?

Source: https://tuoitre.vn/viet-nam-dang-chuyen-dich-sang-ai-20250307081510146.htm