SK hynix is on the cusp of a major advancement in AI memory technology. The South Korean memory giant has announced that it will begin mass production of its new 12-layer HBM3E memory chips by the end of September 2024, with shipments scheduled to commence in the fourth quarter of the same year. This groundbreaking development positions SK hynix as a key player in the rapidly evolving field of AI computing.
Furthermore, SK hynix is already looking ahead to the future. The company has revealed that its next-generation HBM4 memory will be released in the second half of 2025, designed to be fully compatible with NVIDIA’s upcoming Rubin R100 AI GPU. This strategic move underlines SK hynix’s commitment to being at the forefront of AI memory innovation.
The news about the 12-layer HBM3E and the upcoming HBM4 was shared during SEMICON Taiwan 2024, where SK hynix showcased its latest advancements. The company had previously unveiled its 12-layer HBM3E AI memory at FMS 2024 in July, indicating the rapid progress being made. In fact, SK hynix provided NVIDIA with its new 12-layer HBM3E memory modules for testing as early as March 2023. This early testing and quick transition to mass production highlight the company’s dedication to delivering cutting-edge memory solutions.
SK hynix’s ambitions extend beyond merely producing high-performance memory. The company is actively pursuing its goal of becoming the “total AI memory provider” for AI GPUs, a position it is well on its way to achieving through its continuous innovation and unwavering commitment to the field. The upcoming release of HBM3E and HBM4 memory chips will undoubtedly solidify SK hynix’s place as a leader in the AI memory market.