Tens of billions for SK Hynix to stay ahead in AI memory race

Tens of billions for SK Hynix to stay ahead in AI memory race

South Korea’s SK Group is investing a total of $76.4 billion (€71 billion) in its subsidiary SK Hynix over the next three years. This investment is intended to ramp up HBM production and other AI technologies. An additional amount is also set aside for SK Hynix’s repayment of dividends to shareholders.

SK Group, the parent company of chip producer SK Hynix, among others, will invest heavily in its chip production through 2026. This is one of the results of a recent company strategy meeting, Reuters reports.

The strategy meeting was called due to the disappointing results of the company’s two main money makers: chip maker SK Hynix and automotive lithium battery maker SKie Technology. According to top executive Chey Tae-Won, it is necessary for the company to set a new business course to keep up with the rest of the competition.

The emergency consultation resulted in SK Group making a major commitment to its chip production and related technology over the next three years. Through 2026, SK Hynix’s parent company is investing as much as $74.6 billion in its chip production. In addition, another $58 billion will be set aside for developing AI technology and paying dividends to shareholders.

Furthermore, SK Group will further streamline its portfolio of 175 different companies, most of which fall under one of the various SK brands.

SK Hynix gets extra money

The additional $58 million will mainly expand SK Hynix’s production capacity. The investments now pledged focus not only on the development of HBM memory but also, for example, on setting up new data centers and AI services such as GenAI assistants.

All investments are in addition to a previously pledged $90 billion investment for the construction of a mega factory complex in the South Korean city of Gyeonggi.

HBM important in GPUs

SK Hynix is one of the few memory manufacturers worldwide that can design and build high-performance memory for advanced AI workloads. Its direct competitor is American Micron Technologies. Samsung is also a major player, although it recently appeared to be struggling with Nvidia’s requirements.