News Startups

AMD Acquires MEXT to Advance Memory Optimization

AMD acquires startup MEXT, a pioneer in predictive memory technology, in a move to strengthen its AI and data center portfolio. The acquisition expands AMD’s AI capabilities and aims to help customers optimize memory usage, reduce total cost of ownership, and accelerate deployment of high-performance workloads across cloud and enterprise environments.

Modern data center infrastructure is evolving rapidly, with memory emerging as a key constraint as AI models, data analytics, virtualization, and high-performance computing workloads continue to scale in size and complexity. Addressing these bottlenecks has become critical for organizations seeking improved performance per dollar, higher efficiency, and faster deployment at scale.

MEXT has developed AI-driven predictive memory technology designed to make flash storage behave more like DRAM, effectively increasing usable memory capacity while maintaining performance and efficiency. The approach is expected to help reduce infrastructure costs, improve resource utilization, and enable customers to scale both general-purpose and AI workloads more effectively.

The acquisition strengthens AMD’s ability to deliver differentiated, full-stack compute and AI solutions. By integrating MEXT’s technology across its data center portfolio, AMD expects to help enterprise customers unlock greater value from existing infrastructure investments while accelerating AI adoption.

MEXT also brings a specialized team with deep expertise in memory systems and AI infrastructure, which AMD says will further enhance its efforts to address growing data center challenges.

With demand for memory continuing to rise across all categories of enterprise computing, AMD said the combination of its high-performance computing leadership and MEXT’s predictive memory innovation marks an important step toward more efficient, scalable, and cost-effective AI and data center deployments.