Siam.ai Cloud CEO Ratanaphon Wongnapachant celebrated the company’s achievements as one of the first countries in Southeast Asia to receive world-class infrastructure for technologies such as the NVIDIA GB200 NVL72.
The acquisition reflects Thailand’s world-leading situation in creating AI infrastructure. After revealing this supercomputer module, the company hopes to receive more demand from domestic and foreign users.
Through AI capabilities and advanced processes, the NVIDIA GB200 NVL72 accelerates innovation in healthcare, finance, manufacturing, smart cities and other industries for businesses, research and government agencies.
This supercomputer module costs around 100 million. Two of the new GPUs are owned by SIAM AI with the aim of importing an additional 46 by May this year. However, its server equipment costs around 1.55 billion people, and also has a network and installation process, which costs around 200 million people.
As for the Nvidia DGX Blackwell B200, Siam.Ai Cloud had already bought eight and had 1.4TB of GPU memory space. Each cost THB11 million, but it can be three times faster than its predecessor model and 15 times faster than the current model used to train Siamgpt. siam.ai Cloud plans to release these supercomputer modules by 2Q25.
Siam.ai Cloud will become Thailand’s first company to be selected as an NVIDIA Cloud Partner (NCP), and new AI cloud infrastructure will be built in Chon Buri and Pathum Thani with an investment of THB32.5 billion. Previously, the company received investment support from the Investment Committee (BOI) through tax exemptions. This project is one of many projects that the government can support and start attending customers, mostly large hyperscale organizations.
Additionally, there is the acquisition of GPU clusters worth more than THB7 billion. The company predicted that its investment in AI infrastructure would exceed THB70 billion this year, with the exception of investments in SIAM GPT and Siam Robotic.
