Micron 128GB DDR5-5600 RDIMM 2Rx4

The Micron 128GB DDR5-5600 RDIMM (2Rx4) represents the cutting edge of server memory technology, leveraging Micron's advanced 1β (1-beta) DRAM node to pack 128 gigabytes of capacity into a single dual-rank, x4-width registered DIMM. Operating at DDR5-5600 MT/s with native on-die ECC plus host-side ECC, the module delivers exceptional data integrity alongside a substantial bandwidth leap over DDR4 generations—providing up to 44.8 GB/s of peak theoretical throughput per DIMM. The 2Rx4 configuration balances rank density with signal integrity, making it compatible with multi-DIMM-per-channel configurations on flagship Intel Xeon Scalable (Sapphire Rapids, Emerald Rapids) and AMD EPYC (Genoa, Bergamo, Turin) platforms.

At 128GB per slot, this RDIMM enables unprecedented memory footprints without requiring exotic buffering technologies found in LRDIMM designs, maintaining lower latency characteristics while still supporting up to several terabytes of total system memory in high-socket-count platforms. The module's Power Management IC (PMIC) adheres to JEDEC DDR5 specifications and supports fine-grained voltage regulation at 1.1V VDD, contributing to improved energy efficiency per gigabyte compared to DDR4 equivalents despite the raw capacity increase. On-module SPD Hub (SPH) functionality provides enhanced BIOS configurability and advanced telemetry reporting to system management controllers.

Launched in Q2 2025, this module targets enterprise workloads that are acutely memory-capacity-bound: real-time analytics engines, large language model (LLM) inference serving, SAP HANA and Oracle In-Memory deployments, and dense virtual machine or container host configurations. Data center architects can populate 24-DIMM platforms with 3TB of system memory per 2-socket server, unlocking configurations previously reserved for purpose-built in-memory appliances. Micron's manufacturing pedigree and rigorous validation against major server platforms ensure broad compatibility across leading server ecosystems.

Ideal for

• In-memory database acceleration: Deploy SAP HANA, Oracle TimesTen, or Redis Enterprise on a fully populated dual-socket server with up to 3TB of addressable DRAM, eliminating storage-tier latency for transactional analytics.
• LLM and AI inference serving: Load large language models or multi-modal AI inference graphs entirely into DRAM across high-core-count EPYC or Xeon platforms, minimizing model-swap overhead and maximizing token throughput.
• Dense virtual machine consolidation: Host hundreds of memory-intensive VMs per host node, enabling server ratio improvements that reduce physical infrastructure footprint in enterprise private cloud environments.
• High-performance computing (HPC) workloads: Support memory-bound scientific simulations, genomics pipelines, and computational fluid dynamics applications that require both high bandwidth and massive working-set capacity.
• Real-time streaming analytics: Power Apache Flink, Apache Spark, or proprietary CEP engines processing high-velocity event streams where all working data must reside in main memory to meet sub-millisecond SLA requirements.
• VDI and DaaS infrastructure: Maximize concurrent virtual desktop sessions per host with abundant per-VM memory allocations, improving user experience density and reducing the total number of physical nodes required.

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.