XCENA Develops CXL Memory Solutions with Integrated Compute for AI Applications

XCENA, founded by former SK Hynix and Samsung engineers, aims to address AI infrastructure challenges by enhancing memory efficiency using CXL technology. The company has raised KRW 68.5 billion (~$50 million) to develop its InfiniteMemory architecture, which combines memory expansion with near-data processing to reduce data movement. XCENA's prototype, MX1, integrates RISC-V compute to optimize memory utilization across systems. The company is focused on scaling customer deployments and preparing for future product generations as the demand for memory-centric solutions grows.

Published Mar 27, 2026

XCENA, established by a team from SK Hynix and Samsung, targets AI infrastructure bottlenecks through advanced CXL memory solutions. The company has raised a total of KRW 68.5 billion (~$50 million) since its inception, with a Series A funding round in 2024 of KRW 60 billion (~$44 million).

AI-curated by Theia

XCENA's InfiniteMemory architecture pairs memory expansion with near-data processing, enabling efficient data handling for AI workloads. The MX1 prototype features embedded RISC-V compute, allowing direct data processing, which reduces data movement. XCENA aims to scale customer deployments and advance next-generation memory products in response to increasing AI demands.