Ingram Micro, a US-based technology distributor, is joining with Taiwanese tech firm Asus and chip maker Nvidia to capitalise on artificial intelligence (AI) investment in the enterprise sector and AI-based data centres, as outlays in these fields is estimated to exceed 10 billion baht per year in Thailand.
Thailand can take advantage of the US-China tech war by attracting investment from foreign firms seeking to relocate their factories, said Pornthep Watcharaamnouy, managing director of Ingram Micro (Thailand).
The country can also become an alternative location for AI-based data centres, competing with Indonesia and Malaysia, he said.
"We estimate the investment in AI cloud-based data centres in Thailand will be at least 10 billion baht per year from foreign operators, while local enterprises are eager to invest, but reluctant because of a lack of use cases for AI and hesitancy on return on investment," said Mr Pornthep.
He cited a survey by the National Electronics and Computer Technology Center (Nectec) that says 71.8% of organisations have the potential to use AI.
Mr Pornthep said the economic challenge has impacted investment in tech spending, but the rise of AI drives demand among corporate users to purchase or refresh their IT systems.
"The fiscal budget to be rolled out in the second half should also stimulate IT spending," he said.
In 2023, a budget of 7.3 billion baht was allocated for AI initiatives by 68 government agencies, up from 6.4 billion in 2022.
The projects focused on human capacity development, infrastructure and data support, R&D in technology and regulations to prepare for widespread use of AI technology.
The rise in demand for generative AI drives the need for computing power for AI applications, which drives demand for graphics processing unit (GPU) servers to perform the tasks needed for each use case, including deep learning and machine learning AI.
Ingram Micro announced a partnership with Asus and Nvidia to deliver generative AI technology, presenting platforms and advanced processing chips assembled into GPU servers that answer the hyper computing needs of organisations.
"We see opportunity from universities, government, financial services, and manufacturing to embrace AI," said Mr Pornthep.
Ingram Micro (Thailand) expects to gain 70% of its revenue from the enterprise market and 30% from the consumer market this year, due to its transformation towards becoming an AI tech provider, representing a shift from a 50:50 ratio of revenue from the corporate and consumer markets previously.
Jason Chung, regional director for East Asia of Asus Global Pte, said Asus is focusing more on high-performance computing (HPC) data centres by providing GPU servers equipped with advanced air cooling and liquid cooling solution technology. These enhancements ensure superior performance and efficiency, he said.
"Enterprises will need GPU servers to leverage their specific use cases of their large language model," said Mr Chung.
Peter Wu, Asus Cloud and president of Taiwan Web Service Corporation (TWSC), said TWSC, a subsidiary of the Asus Group, plans to invest in Thailand for a cloud-based data centre.
"We have a public-private partnership [PPP] model that is successful in Taiwan and we aim to bring the model to Thailand through a joint venture or other models to invest in AI HPCs with a research institution such as Nectec, or another local enterprise to lower barriers for AI computing in Thailand. This includes large enterprises, government and small enterprises," said Mr Wu.
Thailand is in a good position for AI data centres as it has a large population, locations as well as a large language model for the Thai language.
"Thailand can support Indochina countries for the processing and building of their large language models for their local languages," said Mr Wu.