Samsung Electronics Co., Ltd. recently unveiled its latest innovations in AI memory at the Future of Memory and Storage (FMS) 2026 event in Santa Clara, California.

With extensive expertise spanning DRAM, NAND, enterprise storage and related advanced semiconductor technologies, Samsung has unveiled its latest AI memory portfolio and technology roadmap. This includes an initial unveiling of the zHBM and zNAND-O concept models, the industry’s first demonstration of V10 BV-NAND with more than 400 layers, as well as a comprehensive range of solutions designed to strengthen future AI infrastructure. Samsung also presented its stand at FMS 2026, inspired by an AI cloud server, where around 30 memory and storage technologies were on display.

At the opening keynote event, executives from Samsung’s Memory Business – Jin-Yub Lee, Executive Vice President and Head of Flash Product & Technology, and Kyungryun Kim, Vice President and Project Leader of the DRAM Design team – spoke about next-generation 3D structures in future memory solutions.

Entitled “Driving the Wave of the AI Revolution: 3D Innovations in Memory & Storage Architecture”, the keynote presented Samsung’s vision for maximising the performance and energy efficiency of AI systems, whilst meeting the rapidly growing demands of high-performance computing (HPC) and AI infrastructure.

The company also unveiled the V10 BV-NAND, an architecture that utilises new wafer bonding technology, a first for the industry.

Samsung’s vision for next-generation 3D memory

At FMS 2026, Samsung unveiled the concept models of zHBM and zNAND-O for the first time in the industry, revealing its vision for the next generation of 3D memory architectures.

Designed for advanced AI computing, zHBM introduces a new memory architecture that stacks HBM vertically directly above the AI accelerators, surpassing conventional designs where HBM is positioned alongside the processor.

By minimising the distance that data travels between the AI processor and the memory, zHBM is designed to offer higher bandwidth and improved energy efficiency, enabling the extremely fast data processing required for large-scale AI training and inference.

A next-generation interface system incorporating zHBM is expected to offer approximately eight times the performance of HBM5. With next-generation wafer bonding technology, zHBM can achieve more than 10 times the memory density of HBM5, whilst improving energy efficiency threefold and reducing thermal resistance by more than half. This maximises the stability and energy efficiency of high-capacity, high-bandwidth systems.

zHBM also supports designs tailored to customers’ needs, enabling the integration of customised IP at the intermediate layer between the memory and the AI accelerator, with the aim of expanding memory capacity and enhancing the accelerator’s performance.

Building on its close collaboration with its customers, Samsung plans to further optimise the integration of AI processors and memory via zHBM, in order to maximise the performance of AI systems.

The company also unveiled zNAND-O, a next-generation, high-performance NAND solution based on Samsung’s V-NAND technology, which is available in four- and eight-layer versions. Combining high spatial efficiency, improved I/O performance and low latency, zNAND-O is optimised for edge AI environments that support real-time, data-intensive AI applications.

Industry-leading V10 BV-NAND technology

Samsung has unveiled its V10 BV-NAND technology, which features the new Bonding V-NAND architecture. The announcement comes 13 years after Samsung unveiled the industry’s first V-NAND technology at the Flash Memory Summit 2013, marking yet another milestone in the company’s innovation in the NAND sector.

With more than 400 layers, the V10 BV-NAND offers higher performance and energy efficiency, whilst significantly increasing storage density. By applying wafer bonding technology to stack the memory cells, Samsung has increased memory density by approximately 58 per cent compared with the previous generation, the V9.

In addition to the increased number of layers, the V10 BV-NAND also improves read, write and I/O performance compared with the previous generation, supporting high-performance, high-capacity NAND for future AI systems and their applications.

AI memory roadmap from HBM to PIM and enterprise storage

Samsung also unveiled its latest portfolio of AI memory and storage solutions, including HBM4E, HBM5, LPDDR5X-PIM and PM1763, highlighting its roadmap for AI computing and next-generation data centre infrastructure.

Following the industry’s first mass production of HBM4, using 1c DRAM and 4-nanometre (nm) base die technologies in February, Samsung became the first company to begin shipping HBM4E samples to customers worldwide in May, reinforcing its leading position in next-generation HBM technology.

Alongside the HBM4E samples and the HBM5 prototype, the company unveiled LPDDR5X-PIM, the industry’s first LPDDR memory with processing-in-memory technology. The new solution is designed to perform data processing within the memory itself, improving efficiency in both data transfer and energy consumption.

The exhibition also showcased Samsung’s latest enterprise storage solutions, including the PM1763 and BM1773, designed to meet the growing storage needs of AI data centres.

One-stop turnkey solutions

As an integrated device manufacturer (IDM) with capabilities spanning memory, foundry and advanced packaging, Samsung has unveiled its end-to-end strategy for next-generation AI infrastructure. This strategy offers customers one-stop turnkey solutions that support the evolving needs of the AI era.

From product design through mass production, Samsung delivers integrated development and manufacturing services that help shorten development cycles, while optimising performance and power efficiency, enabling customers to quickly bring differentiated AI semiconductor solutions to market.