Key Takeaways
- Nvidia selected Samsung and SK Hynix exclusively to provide HBM4 memory for Vera Rubin, its next-generation AI accelerator
- Micron was excluded from the Vera Rubin flagship supply chain, causing MU shares to decline 6.74%
- Samsung successfully met Nvidia’s HBM4 quality standards at 10 Gbps and 11 Gbps speeds; SK Hynix continues working toward the 11 Gbps benchmark
- SK Hynix will deliver more than 50% of Nvidia’s HBM requirements in 2026; Samsung’s portion increases to 28%
- Manufacturing is scheduled to commence in March, with Vera Rubin expected to debut during the latter half of 2026
According to Korea Economic Daily, Nvidia has designated Samsung and SK Hynix as the sole providers of sixth-generation high-bandwidth memory (HBM4) for its forthcoming Vera Rubin AI accelerator. Notably absent from this flagship chip cycle is Micron, which previously served as an important HBM supplier.
The announcement triggered a 6.74% drop in Micron’s stock price. Samsung’s shares on the Korean exchange declined 7.81%, while SK Hynix tumbled 9.52%. Nvidia experienced a 3.01% decrease.
Vera Rubin represents Nvidia’s upcoming flagship AI platform, following the Blackwell architecture. The complete NVL72 rack setup combines 72 Rubin GPUs with 36 Vera CPUs, achieving 10x superior performance per watt compared to Blackwell.
Micron isn’t completely excluded from the picture. The company will provide HBM4 for Rubin CPX, a mid-range inference-oriented accelerator within the Rubin family. However, it won’t participate in the premium Vera Rubin offering.
Samsung received approval after successfully meeting Nvidia’s internal quality standards at both 10 Gbps and 11 Gbps performance levels. SK Hynix continues efforts to achieve the 11 Gbps threshold but maintains its position as the leading HBM supplier globally.
SK Hynix Maintains Dominance While Samsung Expands
SK Hynix is anticipated to control approximately 50% of worldwide HBM production in 2026, representing a modest decline from 59% in 2025. Samsung is positioned to increase its market share to 28%, rising from 20% the previous year.
SK Hynix is forecasted to deliver over half of Nvidia’s combined HBM needs — including HBM3E — throughout 2026, and is anticipated to dominate Vera Rubin HBM4 volume.
Both manufacturers are expected to initiate HBM4 production this month. Vera Rubin remains scheduled for a second-half 2026 release.
Vera Rubin’s Intended Applications
The Vera Rubin platform targets large-scale AI training and inference workloads, including the increasingly prevalent mixture-of-experts (MoE) models in cutting-edge AI development.
Reported potential clients include Microsoft, Amazon, Oracle, and Google. These represent the same cloud hyperscalers that have constituted Nvidia’s largest customer base in recent years.
The Vera Rubin NVL72 consumes twice the power of Blackwell but provides significantly enhanced output per watt, a critical factor for data center operators managing these systems at scale.
HBM4 delivers increased memory bandwidth compared to earlier generations, addressing one of the primary bottlenecks in training and operating large AI models.
Wall Street analysts maintain bullish views on Nvidia. Based on TipRanks data, NVDA holds a Strong Buy rating from 39 analysts, with one Hold recommendation. The consensus price target of $272.16 suggests approximately 53% potential upside from present levels.
Throughout the past year, Nvidia stock has appreciated 66.2%, despite Monday’s retreat following the supplier announcement.
Samsung and SK Hynix are positioned to launch HBM4 production in March 2026, with Vera Rubin hardware anticipated to reach customers during the year’s second half.
