[Edaily Reporter SOYEON KIM ] SamsungElectronics and SK hynix will both participate in “FMS (Flash Memory Summit) 2026,” the world’s largest conference dedicated to memory and storage. The event brings together global AI and semiconductor companies, including NVIDIA, to unveil their latest memory technologies and strategies.

According to industry sources on the 14th, FMS 2026 will be held from the 4th to the 6th of next month (local time) at the Santa Clara Convention Center in California, U.S. While the event previously focused on NAND flash, it has expanded into a global gathering covering the entire memory sector—including DRAM, Compute Express Link (CXL), and next-generation storage—as the era of artificial intelligence (AI) gains momentum.

Celebrating its 20th anniversary this year, the event is a leading global semiconductor conference that brings together experts and companies in the fields of AI, machine learning (ML), data analytics, and high-performance computing (HPC).

This year’s event will feature a large-scale unveiling of cutting-edge memory and storage technologies for data centers and AI infrastructure. In addition to NVIDIA, companies such as Marvell, Intel, Micron, Microsoft (MS), SanDisk, Kioxia, and Solidigm are scheduled to participate and share the latest technology trends.

On the first day of the event, April 4, SamsungElectronics Executive Vice Presidents Song Taek-sang and Lee Jin-yeop will deliver keynote speeches. They are scheduled to introduce next-generation memory and storage strategies required for the AI era.

As AI systems perform tasks such as planning, inference, long-term memory retention, and multi-agent collaboration, the demand for memory performance and storage capacity is rapidly increasing. Consequently, the importance of next-generation memory technologies capable of resolving bottlenecks in AI infrastructure is becoming increasingly prominent.

SamsungElectronics will present its AI infrastructure strategy centered on its next-generation memory and flash storage portfolio. Leveraging its capabilities as an integrated device manufacturer (IDM)—encompassing memory, foundry (semiconductor contract manufacturing), and advanced packaging—the company plans to introduce integrated solutions that can reduce bottlenecks in AI infrastructure and enhance scalability.

At SK hynix, Executive Vice Presidents Kim Chun-sung and Kang Wook-sung will deliver keynote speeches. Under the theme “Efficient AI Infrastructure Orchestration via Tiered Memory in the Era of Agent AI,” they will present strategies for building efficient AI infrastructure. They will propose tiered memory systems as a key strategy for resolving data bottlenecks. They explain that memory tiers can be implemented using technologies such as 3D-stacked DRAM for xPUs, high-bandwidth flash (HBF), and CXL-based memory pooling.

The company will also unveil its next-generation storage strategy. As the proliferation of Agent AI leads to a surge in KV cache data, increasing the importance of high-capacity, high-performance storage, SK Hynix plans to present solutions that simultaneously ensure cost-efficiency and storage capacity through technologies such as QLC-based SSDs.

Engineers from SamsungElectronics and SK hynix will also lead various sessions on key memory technologies for the AI era, such as CXL and HBF. Jo Jeong-hyun, a technology leader at SK hynix, will explain how HBF reduces bottlenecks in data centers and outline future directions for the technology. Micron will also unveil its CXL-based distributed memory technology. The company plans to explain how CXL resolves memory capacity bottlenecks in large-scale AI inference environments and expands the effective memory bandwidth available for KV caching.

An industry official stated, “Demand for memory to support next-generation AI infrastructure continues to grow,” adding, “This event will serve as an opportunity to gain insight into the direction of next-generation memory technology.”