Nvidia is accelerating its push into the data center market with its next-generation central processing unit (CPU) “Vera,” fueling growing expectations of a windfall for South Korea’s leading memory companies, Samsung Electronics (005930.KS) and SK Hynix (000660.KS). As AI infrastructure evolves from GPU-centric systems to integrated platforms combining CPUs, memory, and storage, Nvidia’s strategic expansion is directly translating into supply opportunities for the South Korean memory industry.

According to industry sources on the 11th, Nvidia’s new Vera CPU is positioning itself as a core component of the next-generation AI platform “Vera Rubin.” Since this platform incorporates sixth-generation high-bandwidth memory (HBM4), low-power server memory, and high-performance storage, the prevailing view is that wider platform adoption will drive proportional growth in the supply scale of related components.

Within the HBM supply chain in particular, Samsung Electronics and SK Hynix are regarded as key Nvidia partners. During a visit to South Korea last month, Nvidia CEO Jensen Huang stated that “major memory companies including Samsung Electronics, SK Hynix, and Micron have passed HBM4 qualification, and competition is underway for Vera Rubin platform supply.” This signals that both South Korean memory giants have secured their positions in the core supply chain for the next-generation AI platform.

Samsung Electronics is actively expanding its supply scope beyond memory into the storage sector. At the GTC 2026 event in March, the company showcased HBM4 for Rubin GPUs, SOCAMM2 for Vera CPUs, and the enterprise SSD PM1763 alongside the Vera Rubin platform. The recently mass-produced PCIe 6.0-based eSSD PM1763 was introduced as a core storage device for Vera Rubin. As data volumes surge during AI model training and inference, the importance of high-performance storage that reduces GPU and CPU latency is coming to the fore. Samsung’s strategy is to further solidify its position within Nvidia’s next-generation platform by leveraging its capability to supply HBM4, SOCAMM2, and eSSDs as a package.

SK Hynix is also moving swiftly to respond to the Vera Rubin ecosystem. The company has begun mass production of a 192GB SOCAMM2 product based on 10nm-class 6th-generation (1c) LPDDR5X. SOCAMM2 is a memory module that re-engineers mobile low-power DRAM for server environments, characterized by low power consumption, high bandwidth, and a detachable structure. Optimized for Nvidia’s Vera Rubin platform, it demonstrates SK Hynix’s competitiveness in the next-generation server memory market.

With the Vera CPU, Nvidia is making a full-fledged entry into the server CPU market long dominated by Intel and AMD. Nvidia is reportedly expecting approximately $20 billion (around 30 trillion won) in revenue from the Vera CPU business by the end of this fiscal year. This aggressive target appears to be materializing through strong interest from major AI companies. AI search startup Perplexity recently announced plans to adopt Vera CPUs, following earlier disclosures from OpenAI, Anthropic, and Oracle regarding their intention to utilize Nvidia CPUs.

Samsung Electronics is showing signs of seeking to extend its partnership with Nvidia beyond memory and storage into the foundry (semiconductor contract manufacturing) space. Jun Young-hyun, Vice Chairman and head of Samsung Electronics’ Device Solutions (DS) Division, met with CEO Jensen Huang last month to discuss next-generation memory development including HBM4E and HBM5, as well as foundry collaboration. The two companies are reportedly also cooperating in the production of autonomous driving chips and Grok AI accelerator chips.

Industry observers anticipate that the proliferation of AI agents will trigger full-scale competition among Big Tech companies to develop their own CPUs. “The spread of AI agents is sparking CPU development competition among Big Tech,” one industry insider noted, adding that “as more AI platforms emerge that optimize CPUs, GPUs, and memory together, supply opportunities for South Korean memory companies could grow further.” Attention is focused on whether Nvidia’s entry into the CPU market will go beyond a mere shift in competitive dynamics to generate a trickle-down effect across the memory and storage markets where South Korea holds strengths.