Samsung Unveils AI Memory Technology With Next-Generation BV-NAND
Samsung Electronics has introduced a new generation of artificial intelligence memory technology as it seeks to strengthen its position in the rapidly expanding AI semiconductor market. The company unveiled its latest innovations at the Future of Memory and Storage (FMS) conference in Santa Clara, California, highlighting advances designed to deliver greater storage capacity, improved performance and higher energy efficiency for AI applications.
The world’s largest memory chipmaker presented a prototype of its V10 Bonding V-NAND (BV-NAND), featuring more than 400 layers and a new wafer-bonding architecture that significantly increases storage density.
Samsung Introduces V10 BV-NAND
Samsung said its new BV-NAND technology improves memory density by approximately 58% compared with its previous V9 NAND generation.
The company added that the technology also enhances read, write and input/output performance to meet rising demand from AI systems that require higher-capacity and more power-efficient storage.
As artificial intelligence workloads continue to expand beyond training large language models to generating responses for users, demand for high-capacity NAND flash memory has grown rapidly. NAND memory stores data such as applications, photos and videos across devices including smartphones and other electronics.
New Vision for AI Memory
Alongside the BV-NAND prototype, Samsung showcased concept models of what it described as the industry’s first zHBM and zNAND-O architecture.
The company said these technologies represent its vision for next-generation three-dimensional memory by stacking memory cells vertically to increase storage capacity while improving overall efficiency.
Unlike conventional high-bandwidth memory, which sits beside AI processors, Samsung’s proposed zHBM architecture places memory directly above AI accelerators. The shorter data path is designed to increase bandwidth while reducing power consumption.
According to Samsung, its wafer-bonding technology could ultimately deliver more than ten times the memory density of conventional HBM5 memory. The company also said the approach could triple energy efficiency and reduce thermal resistance by more than half.
AI Demand Continues to Drive Memory Market
Samsung’s latest announcements come as artificial intelligence remains the primary growth driver for the global memory industry.
Although some investors have expressed concerns that weaker Chinese smartphone demand could affect long-term memory chip sales, analysts continue to point to resilient AI-related demand.
The company’s confidence in the market was reinforced by its disclosure last week that long-term customer agreements could eventually account for between 60% and 70% of its total memory sales.
Analysts Highlight Strong Market Fundamentals
Market analysts also remain optimistic about the outlook for memory chips.
Citi analysts said in a research note last month that inventories across both DRAM and NAND supply chains remain well below historical averages despite concerns about softer demand in some end markets.
The combination of low inventories and sustained investment in artificial intelligence infrastructure continues to support demand for advanced memory technologies, positioning innovations such as Samsung’s BV-NAND for future AI workloads.
With inputs from Reuters