--- title: "SK hynix Unveils Memory-Centric Servers: 256GB CMM-DDR5 Modules, MARS Rack, and CMM-Ax Technology" description: "At the OCP Global Summit in San Jose, SK hynix showcased 256GB CMM-DDR5 modules with speeds up to 80 GB/s, the MARS rack system, and CMM-Ax technology for memory-centric AI servers." date: 2026-10-18T10:48:00.320Z lang: en url: https://xab.info/en/posts/sk-hynix-memory-centric-servers-ocp-global-summit tags: [sk-hynix, cmm-ddr5, cxl, ai-memory, ocp, mars, naver-cloud] author: "Alexander Buran" publisher: "XAB.info" --- # SK hynix Unveils Memory-Centric Servers: 256GB CMM-DDR5 Modules, MARS Rack, and CMM-Ax Technology ![SK hynix exhibition booth showcasing CMM-DDR5 memory and MARS server systems at a conference](https://xab.info/media/2026/10/18/sk-hynix-pamjat-tsentrichnye-servery-ocp-global-summit/sk-hynix-pamjat-tsentrichnye-servery-ocp-global-summit-1.webp) ## Key Points - SK hynix at the OCP Global Summit (October 12–15, San Jose) showcased memory-centric server technologies: CMM-DDR5, MARS, and CMM-Ax. - 2nd-generation CMM-DDR5 modules: 256GB capacity, CXL 3.2 interface, up to 80 GB/s bandwidth, ~30% energy efficiency gain, optimized for pooled memory. - The MARS rack system establishes memory as the central rack resource; developed with Naver Cloud to show shared KV-cache and embeddings for AI agents. - CMM-Ax brings computing closer to memory and was demonstrated running long-context LLMs. - The company emphasizes these are system demonstrations without independent benchmarks, pricing, or release dates. South Korean memory manufacturer SK hynix announced on October 16 that it showcased memory-centric server technologies at the OCP Global Summit in San Jose. The Open Compute Project event—a community developing open hardware for data centers—took place from October 12 to 15. At the core of the company's focus was the concept that memory ceases to be an auxiliary component and becomes the primary resource of a compute rack tailored for artificial intelligence workloads. Among the demonstrations were CMM-DDR5 modules, the MARS rack system, and CMM-Ax technology, which combines computing with memory. ### CMM-DDR5: 256GB and up to 80 GB/s via CXL 3.2 The second-generation CMM-DDR5 module features a capacity of 256GB and operates on the CXL 3.2 interface (Compute Express Link—an open standard for connecting memory and accelerators). According to the company, the bandwidth of these modules reaches 80 GB/s, which is more than double the previous generation, while energy efficiency has improved by approximately 30%. The modules are designed for pooled memory, allowing multiple servers to utilize a shared capacity pool for flexible memory allocation across rack nodes. ### MARS: A Rack Where Memory is the Central Resource The company also introduced the MARS (Memory Centric AI Rack System) rack system, in which memory acts as the central resource of the rack, integrating partner technologies. Together with Naver Cloud, SK hynix demonstrated a concept of shared Key-Value cache (KV-cache)—intermediate data that accelerates text generation in large language models—for an AI agent environment, as well as shared vector embeddings for a service working with various data types. The booth also featured integration with GPUs and HBM, emphasizing a systemic rather than isolated architectural approach. ### CMM-Ax: Computing Right Next to Memory CMM-Ax technology brings computing closer to memory by reducing data movement. It was demonstrated on the task of serving large language models (LLM) with long contexts, where moving large volumes of data between memory and the accelerator traditionally becomes a bottleneck. Thus, SK hynix is building a lineup of solutions ranging from standalone CMM-DDR5 modules to the MARS rack system and the compute-memory CMM-Ax technology, covering various levels of the data center architecture. ### Caveats: System Demonstrations, Not a Product SK hynix emphasizes that these are system demonstrations rather than a widely available server platform. The announcement lacks independently verified benchmarks, pricing, or market release timelines. The stated characteristics—80 GB/s, 256GB, and a 30% energy efficiency boost—are provided by the company itself. The context reflects broader industry dynamics: memory manufacturers are accelerating the rollout of new solutions due to the surging demands of AI systems, making memory-centric architecture a key competitive battleground for the next generation of data centers. ## Fact-Check & Sources - [SK hynix показала память для ИИ-серверов со скоростью до 80 ГБ/с](https://www.ixbt.com/news/2026/10/18/sk-hynix-pokazala-pamiat-dlia-ii-serverov-so-skorostiu-do-80-gbs.html) - Primary source: full text of SK hynix announcement regarding CMM-DDR5, MARS, CMM-Ax, and the Naver Cloud partnership; figures match the core text. - [SK hynix показала память для ИИ-серверов со скоростью до 80 ГБ/с](https://runet.news/news/84321) - Retelling of the same announcement; facts and figures (256GB, 80 GB/s, +30%, CXL 3.2) align with the primary source. - [SK Hynix первой отправила клиентам образцы HBM4E объёмом 48 ГБ и скоростью 16 Гбит/с](https://overclockers.ru/blog/Global_Chronicles/show/257834/SK-Hynix-pervoj-otpravila-klientam-obrazcy-HBM4E-obemom-48-GB-i-skorost-ju-16-Gbit-s) - Context: another SK hynix development (HBM4E), unrelated to the CMM/MARS announcement; used for industry dynamics. - [Привычная скорость работы серверов уйдёт в прошлое: инженеры нашли способ устранить барьер ...](https://www.pravda.ru/news/technolgies/2410601-ai-memory-innovation/) - Context: general theme of memory bottlenecks during inference; does not contradict the announcement. ## FAQ ### What is CMM-DDR5 and what are its specifications? CMM-DDR5 is a second-generation memory module with a 256GB capacity using the CXL 3.2 interface. According to SK hynix, its bandwidth reaches 80 GB/s (more than double the previous generation), with energy efficiency improved by about 30%. The modules are designed for pooled memory. ### What is the MARS system? MARS (Memory Centric AI Rack System) is a rack architecture where memory serves as the central resource, integrating partner technologies. Partnering with Naver Cloud, SK hynix demonstrated a shared KV-cache and shared embeddings for AI agent environments. ### What is the purpose of CMM-Ax technology? CMM-Ax brings computation closer to memory to reduce data transfer overhead, demonstrated in serving long-context large language models (LLMs). ### Is this a finished product available for purchase? No. SK hynix stresses that these are system demonstrations rather than a commercially available server platform. The announcement includes no independent benchmarks, pricing, or release timelines. ### When and where was the demonstration held? At the OCP Global Summit in San Jose, held from October 12 to 15; the company announced the demonstration on October 16.