New memory boosts data centre efficiency

Advantech has announced the release of the SQRAM CXL 2.0 Type 3 Memory Module. Compute Express Link (CXL) 2.0 is the next evolution in memory technology, delivering memory expansion with a high-speed, low-latency interconnect designed to satisfy the demands of large AI Training and HPC clusters. CXL 2.0 builds on the foundation of the original CXL specification, introducing advanced features such as memory sharing and expansion, allowing more efficient utilisation of resources across heterogeneous computing environments.

Fixed allocations often limit traditional memory architectures, resulting in underutilised resources and bottlenecks in data-intensive workloads. With the E3.S form factor, based on the EDSFF standard, the CXL 2.0 Memory Module overcomes these limitations, allowing for dynamic resource management. This not only enhances performance but reduces costs by maximising existing resources.

CXL memory module operates over the PCIe 5.0 interface. This high-speed connection ensures that data transfer remains rapid and efficient even as memory is expanded across different systems. The PCIe 5.0 interface provides up to 32GT/s per lane, allowing the module to supply the bandwidth required for data-intensive applications. Adding more capacity will result in better performance and increased memory bandwidth without the need for more servers and capital expenses.

CXL 2.0 allows multiple hosts to access a shared memory pool, optimising resource allocation and improving overall system efficiency. Through CXL's memory pooling technology, computing components such as CPUs and accelerators of multiple servers on the same shelf can share memory resources, decreasing resource redundancy and solving the problem of low memory utilisation.

CXL memory module can be added or removed from the system without shutting down the server, permitting for on-the-fly memory expansion. For data centres, this translates into the ability to scale memory resources as required, ensuring optimal performance without disruption.