HPE Unveils Next-Gen Memory Server Designed for Compute-Intensive AI Workloads

Hewlett Packard Enterprise (HPE) has unveiled its latest enterprise memory server, the HPE Compute Scale-up Server 3250, designed to handle demanding business workloads, particularly those associated with compute-heavy tasks and emerging agentic AI applications. Powered by the Intel Xeon 6 processors, this new server can support configurations from four to sixteen sockets, boasting up to 64TB of DDR5 memory, tailored for complex transactional and analytical work.

The 3250 utilizes the HPE Superdome Flex architecture, featuring a high-speed interconnect fabric that allows multiple CPU sockets to function as a single system entity. This design aims to minimize server sprawl and streamline data assembly for intensive workloads, such as those running on SAP HANA, Oracle, and Microsoft SQL Server databases.

Additionally, the server incorporates HPE’s Integrated Lights Out (iLO) management technology, enabling efficient management tasks like server diagnostics, access configuration, and automation enhancements aimed at operational efficiency. Krista Satterthwaite, HPE’s senior vice president and general manager for compute, highlighted the server’s security features, including a silicon root of trust which safeguards against future threats with post-quantum cryptography.

The HPE Compute Scale-up Server 3250 offers advanced memory error detection and correction capabilities, alongside features aimed at memory healing and deconfiguration, enhancing its reliability and uptime for critical business applications. This new system is strategically positioned to compete against IBM Power servers and the premium offerings from industry players like Dell and Fujitsu.

The HPE 3250 is available for purchase now.

Total
0
Shares
Leave a Reply

Your email address will not be published. Required fields are marked *

Previous Article

How CUDA Highlights Nvidia's Evolution into a Software Powerhouse

Related Posts