Samsung Electronics Co. is reportedly nearing a significant milestone, approaching final certification from Nvidia Corp. for its latest generation of artificial intelligence memory chips, known as HBM4. This development marks crucial progress for the South Korean tech giant in its efforts to reduce the competitive lead currently held by rival SK Hynix Inc. The Suwon-based company has entered the ultimate qualification phase with Nvidia, a key customer that utilizes vast quantities of high-bandwidth memory (HBM) for its cutting-edge AI accelerators. Samsung had previously provided initial samples of its HBM4 chips to the US chipmaker in September, indicating a steady progression in its development and testing.

Sources familiar with the matter, who requested anonymity due to the private nature of the information, indicate that Samsung is making preparations for the mass production of its HBM4 chips, with an anticipated start in February. While the company is poised to begin shipping these advanced memory components soon, the precise timing of these shipments remains uncertain. News of Samsung's progress had an immediate impact on the market, with its shares experiencing a gain of up to 3.2% in Seoul on Monday before stabilizing, while SK Hynix's stock saw a corresponding decline of approximately the same margin. A representative for Samsung chose not to comment on these developments.

Despite this positive movement, Samsung currently lags behind industry leaders SK Hynix and Micron Technology Inc. in the forefront of the highly specialized AI memory sector. Nevertheless, all three dominant memory chip manufacturers have witnessed a dramatic surge in their stock values in recent weeks. This surge is directly attributed to the intense global demand for artificial intelligence, which has subsequently led to a significant shortage of memory components across the broader electronics industry. Collectively, these three leading chipmakers have amassed an impressive roughly $900 billion in additional market value since the beginning of September.

Investor optimism is now mounting regarding Samsung's potential to join its competitors in supplying vital components for Nvidia's forthcoming flagship Rubin processors. Historically, Nvidia has predominantly relied on SK Hynix for its most sophisticated memory chips, which are intricately paired with its top-tier AI accelerators. Recent reports from The Korea Economic Daily further bolster this expectation, suggesting that Samsung is slated to commence HBM4 shipments not only to Nvidia but also to Advanced Micro Devices Inc. (AMD) as early as next month. Both Samsung and SK Hynix are scheduled to conduct their respective earnings calls on Thursday, during which they are widely expected to provide updates and discuss the progress made on their HBM4 chip technologies, offering further insights into the evolving landscape of AI memory.