SK Hynix and Marvell Technology jointly unveil CXL memory module architecture for AI data centers

SK Hynix has unveiled a CXL (Compute Express Link) memory module capable of handling computation directly, developed in collaboration with Marvell Technology, a global system chipmaker. The module, called CMM-Ax, represents a CXL-based processing-near-memory (PNM) memory solution positioned to advance AI data center infrastructure.

CXL is widely viewed as the next-generation AI semiconductor to follow HBM. Unlike HBM, which attaches ultra-fast memory directly to a GPU, CXL is an interconnect technology that enables CPUs, GPUs, and memory chips to communicate seamlessly while flexibly connecting and expanding memory across an entire server system. SK Hynix is moving beyond simply expanding memory capacity by integrating computational functions to prepare for the era of AI inference.

The CMM-Ax combines SK Hynix's high-performance DRAM with Marvell's CXL-based intelligent PNM engine, Structera A, enabling more efficient memory use within AI data centers. The CXL controller integrates 16 high-performance data center CPUs from Arm, which enables in-memory computation at bandwidths of up to 200 gigabytes per second.

The collaboration addresses a critical bottleneck in modern AI systems. As AI shifts from training to inference, large language models store previously computed information in a KV cache (Key-Value Cache) to improve efficiency. This eliminates the need to repeat identical computations but causes KV cache capacity to balloon rapidly. The CMM-Ax addresses this by processing KV cache operations within the memory domain rather than on the GPU, significantly reducing data transfer latency.

When applied to a GPU-based backend system called NELSSA, the CMM-Ax achieved throughput up to 5.5 times higher than a single GPU and 3.6 times higher than a dual-GPU configuration, with validation conducted using 512 gigabytes of capacity per device. The companies confirmed the module can deliver high-performance inference in long-context LLM environments.

SK Hynix and Marvell are presenting their CMM-Ax findings together at FMS (Future of Memory and Storage) 2026, a leading global memory technology conference being held in Santa Clara, California. According to Joo Young-pyo, executive vice president at SK Hynix, "Unlocking AI's full potential requires intelligent memory solutions that combine large capacity with high-performance computing. The successful validation of CMM-Ax is the fruit of a powerful synergy between Marvell's system-on-chip expertise and SK Hynix's leadership in AI memory."

The semiconductor industry expects CXL to follow a growth trajectory similar to HBM, with some observers calling it the "next HBM" in the AI chip market. CXL can deliver performance comparable to DRAM while significantly expanding capacity, potentially cutting the cost of building and operating AI data centers. According to market research firm Yole Group, the CXL market is valued at $2.1 billion in 2026 and is projected to grow to approximately $16 billion by 2028.