Micron Unveils SOCAMM Memory Modules for NVIDIA GB300 Series at GTC 2025
March 19, 2025 – At the GTC 2025 conference, Micron Technology officially introduced its SOCAMM (Stacked-On-Carrier Attached Memory Module) solution, which will be integrated into NVIDIA’s newly launched Blackwell Ultra GB300 product line.
March 19, 2025 – At the GTC 2025 conference, Micron Technology officially introduced its SOCAMM (Stacked-On-Carrier Attached Memory Module) solution, which will be integrated into NVIDIA’s newly launched Blackwell Ultra GB300 product line.
Micron confirmed that SOCAMM has now entered mass production, offering unparalleled speed, compact size, low power consumption, and high capacity. This advanced memory module is designed to meet the increasing demands of AI workloads, delivering high-bandwidth, energy-efficient memory for AI servers and data-intensive applications.
SOCAMM vs. RDIMM: A Game-Changer for AI Workloads
Micron’s SOCAMM technology surpasses traditional RDIMM modules with:
2.5x the bandwidth at the same capacity, allowing for faster access to larger AI training datasets and enhanced inference throughput.
Ultra-compact size (14 x 90 mm)—only one-third the size of industry-standard RDIMMs—enabling high-density AI server configurations.
Power efficiency, leveraging LPDDR5X memory, reducing power consumption to just one-third of standard DDR5 RDIMMs—crucial for energy-efficient AI data centers.
Key Features of SOCAMM Memory Modules
128GB capacity, achieved through four 16-die stacked 16Gb LPDDR5X chips.
128-bit memory bus running at 8533 MT/s, accelerating AI model training and supporting higher inference workloads.
Advanced scalability and reliability, with enhanced error correction mechanisms, ensuring data integrity in AI-driven environments.
Optimized for AI Servers and Liquid Cooling
Unlike LPCAMM2, Micron’s SOCAMM modules:
Feature a single-sided four-chip design, secured with three mounting screw holes for improved stability.
Eliminate the trapezoidal top structure, reducing overall height to enhance compatibility with liquid-cooled AI servers.
By integrating SOCAMM into NVIDIA’s Blackwell Ultra GB300 series, Micron is redefining AI memory solutions, delivering a cutting-edge, high-bandwidth, and power-efficient memory architecture for next-generation AI computing and cloud infrastructure.
