Chips are no longer just electronic components hiding inside devices, but are becoming the infrastructure of the AI economy, as global competition moves from developing models to having the ability to operate them efficiently and at scale. As demand for computing chips escalates, the semiconductor industry intersects with energy, data centers, and digital infrastructure, creating a new map of technical power that goes beyond the borders of the countries that manufacture the chips to the countries capable of absorbing the computing they need.
In this race, Saudi Arabia is emerging as a growing destination for Asian chip and artificial intelligence companies, driven by the availability of energy, the rapid expansion of data centers, and the ambition to build a sovereign artificial intelligence ecosystem. In contrast, South Korea is moving to strengthen its position at the heart of the value chain, by increasing government spending on chips, artificial intelligence, and related infrastructure.
The decision of the Korean company “Rebellions”, which specializes in artificial intelligence chips, to choose Riyadh as its regional headquarters embodies this new intersection between those who manufacture chips and those who have the ability to operate them. Its CEO, Sungyeon Park, told Asharq Al-Awsat that three main factors make Saudi Arabia the company’s destination: high energy levels, commitment to developing artificial intelligence infrastructure, and great ambition in the field of sovereign artificial intelligence.
Park added that the combination of these factors makes Riyadh a suitable location for Rebellion to develop its business in the field of artificial intelligence chips.
The Korean company's move comes at a time when Saudi Arabia is working to expand its capacity to accommodate the growing demand for computing, with the growth of investments in data centers and artificial intelligence projects.
This week, South Korea proposed a government budget, the largest in its history, worth 821 trillion won ($596.92 billion) for the year 2027, an increase of 12.8 percent from the current year, in the largest annual increase ever, with a large portion of its spending directed to artificial intelligence, semiconductors, and related infrastructure.
Seoul is betting on increasing public spending to maintain its position in the semiconductor industry, which constitutes one of the most important engines of its economy, in light of the rising global demand for chips used in artificial intelligence applications.
The draft budget includes allocating 21.3 trillion won for industrial water systems, electricity networks, and logistics services needed to support the next generation of semiconductor infrastructure, in addition to 2.6 trillion won in a special budget for chips.
South Korea is benefiting from a global surge in demand for high-bandwidth memory (HBM) chips, used in artificial intelligence systems, with Samsung Electronics and SK Hynix recording strong profits driven by this demand.
Park believes that the sovereign dimension of artificial intelligence has become an important factor in countries' decisions regarding the technologies they use, especially with some data and applications linked to sensitive sectors. He said that the need for locally managed artificial intelligence technologies is related, among other things, to “national security,” noting that some sensitive information, such as nuclear energy or weapons programs, requires the state’s ability to control how artificial intelligence is used, set its own limits, and operate it entirely within the country.
This reflects one of the major trends in the AI market, as governments and companies focus not only on access to advanced models, but also on where data is stored, where models run, and who owns the infrastructure that relies on it.
Rebellion already has an investment connection to Saudi Arabia, as Park said that Aramco invested in the company about two years ago, while Rebellion is in ongoing talks with Humane about strengthening supply chains for artificial intelligence infrastructure technologies.
He said that Humane has become one of the most important institutions working in artificial intelligence in Saudi Arabia and at the global level, and that cooperation with it can leave a positive impact on the artificial intelligence sector in the region.
These moves coincide with Saudi Arabia's expansion into digital infrastructure, including data centers, which provides companies specializing in chips and computing with a potential regional market that goes beyond simply selling components to participating in building the artificial intelligence system itself.
Rebellion does not believe that competition in the artificial intelligence chip market will be decided solely by the chip's ability to process a greater number of codes per second, but also by the cost of operating artificial intelligence in the long term.
Park said that the company is focusing on developing technologies that provide “higher artificial intelligence performance at a lower economic cost,” explaining that the important metric for customers is not limited to the number of codes that can be processed per second, but rather includes “the cost per code.”
The company's model focuses largely on inference processes, rather than distributing effort evenly between inference and model training.
Park explained that heuristics is gradually becoming a commodity, which means companies and users will be more interested in getting good results at a reasonable cost, regardless of the type of technology behind it.
This factor is particularly important with the rise in energy consumption associated with data centers, as the cost of electricity and chip efficiency have become part of the economic calculation for expanding the use of artificial intelligence.
Here the calculations of Riyadh and Seoul meet from two different angles. While South Korea is investing in the chips, electricity grids, and industrial infrastructure needed to maintain its position in the value chain, Saudi Arabia is working to provide the energy, infrastructure, and data centers needed to accommodate the next wave of computing. Thus, the AI race between countries is not only about models and software, but about who makes the chips, who owns the data centers, who can provide energy, and who provides all of this at the lowest possible cost.
