SEOUL, South Korea / RankWire.AI / – Samsung Electronics has unveiled innovative strategies for memory and storage technology designed to enhance artificial intelligence infrastructure. The company shared its plans at the 2026 OCP Global Summit in San Jose, California, held from October 12 to 15. Its roadmap emphasizes high-performance memory, expansion via Compute Express Link, and the development of next-generation solid-state drives. Samsung will also collaborate with AMD on a joint CXL demonstration. These initiatives address current AI server capacity limitations.
Samsung outlined its approach through three main objectives: increasing speed, improving memory scalability, and ensuring data persistence. For speed, it introduced customized high-bandwidth memory along with HBM5, the eighth generation of the technology. To enhance scalability, the company promoted CXL as a method to extend capacity beyond the memory available inside each server. For data persistence, Samsung presented next-generation SSD storage solutions and management systems. It also showcased storage software aimed at boosting data processing efficiency in AI applications.
Samsung highlighted the rising demand for more memory in AI inference tasks. Larger models generate more data during response generation. This data must remain accessible to processors. Traditional systems limit how much memory each processor can access. Samsung presented CXL technology as a practical solution to that challenge. It also demonstrated key-value cache optimization, enabling AI models to reuse processed data instead of repeating work.
Partnership with AMD expands to CXL memory expansion solutions
Samsung and AMD will jointly showcase a CXL-based system that increases memory capacity in AI setups. This technology allows processors to share memory resources, rather than relying solely on local server memory. Samsung reports that engineers have tested the solution in real system environments. The demonstration follows a memorandum of understanding signed by the two companies in March 2026 for next-generation AI memory and computing. AMD CEO Lisa Su visited Korea on October 7 and met Jeon Young-hyun, Vice Chairman of Samsung’s Device Solutions division.
Memory pooling was another focus of Samsung’s presentation. The company demonstrated CMM-D, a CXL-based DRAM module connecting multiple memory units into a shared pool. Systems can draw capacity from this pool as needed. Samsung also exhibited processing-in-memory technology, which performs some calculations inside memory chips to reduce data movement and save energy. Exhibits included 2-nanometer gate-all-around wafers, Groq LPU wafers, and Nvidia’s Groq 3 LPX system.
Early co-design efforts influence Samsung’s AI memory development plans
Shim Ho-jun, Corporate Vice President and Head of Samsung’s Application Engineering Team, shared details during an executive session. He emphasized that larger AI models and increased inference demands require greater memory capacity and bandwidth for optimal system performance. Shim also noted that Samsung is expanding co-design and validation work with global partners from the earliest stages. The goal is to better align its memory and storage products with customer system designs.
Samsung MAX Lab provides clients with a space to evaluate memory and storage solutions under real-world AI workloads. The lab also offers co-design services tailored to specific system configurations. Customers can test pre-commercial technologies, including PIM and next-generation SSDs, before they are commercially available. Samsung participated in the summit as a Diamond Sponsor. This event follows its 2026 appearances at Nvidia GTC, COMPUTEX, and the Future of Memory and Storage conference, as the company seeks to expand its role in AI memory innovation.