Why Does Thailand Need Its Own Data Centers When They Can Be Anywhere In The World?

This question might sound a bit cheeky, but it's worth considering: If foreign countries are better at it, and we have to buy processing chips from them, where does Thailand find an opportunity to compete?

Ajarn Fon, founder of Carbonoi, has some interesting insights to share.

Going back to 2020, Chulalongkorn University was the first in Thailand to invest millions of baht in an NVIDIA DGX A100 machine for research. But soon after, new models kept emerging. Now, government agencies, private companies, and academic institutions are all spending hundreds of millions to buy GPU servers themselves. Once you're in this field, you have to accept that technology moves incredibly fast. Only a few companies in the world produce high-performance chips.

Theoretically, data centers can be located anywhere in the world. We might not realize that a significant amount of Thai data resides on foreign servers in places like the US, India, and Singapore. Cloud rental costs vary by region, differing in speed, efficiency, and submarine cable security. For instance, prices in Singapore are higher than in India.

If the Thai government promotes investment, such as tax incentives, domestic cloud rental costs could decrease. This must be coupled with building credibility in data security, as hackers can attack from anywhere globally.

Once we have the #hardware, the most beneficial highlight for the country will be developing #software tailored to the local context, especially Thai-language Large Language Models (LLMs), Optical Character Recognition (OCR) for reading text from images, and Speech-to-Text converters. We can do these better than foreign solutions.

Currently, Thai researchers are developing innovations using Deep Learning, spread across universities and the private sector, including banks, insurance, retail, and public health. This is because they're frustrated by two major problems when using foreign services:

It's expensive! And most cloud services offer no support when problems arise!

This has led many organizations to start looking for Thai service providers who understand Thai problems better. In the long run, having our own systems with domestic databases offers more advantages than disadvantages, both in terms of data security and controllable pricing.

The AI market is moving very fast, with providers everywhere you look. But I want everyone to look beyond just technology, towards aspects like winning over customers with continuous quality and excellent service. Thais already have an advantage here 😉. Even if data centers aren't setting up bases in Thailand rapidly, other tasks can be done on computers without borders, just as Carbonoi is developing AI for Climate Change tailored for Thailand.