Super Micro Computer has rolled out its latest AI infrastructure lineup, introducing the H15 server portfolio powered by AMD's 6th Gen EPYC 9006 processors. These new servers target enterprise AI, cloud computing, and high-performance workloads, boosting processing power while keeping energy consumption stable.
Expanded Capabilities for AI and Cloud Workloads
The H15 series supports up to 256 processor cores and 512 threads, enhancing memory capacity and increasing input/output bandwidth. The lineup includes multiple platforms like Hyper, CloudDC, GrandTwin, FlexTwin, Petascale Storage, and SuperBlade, catering to use cases from cloud providers to research centers. Customers can also choose between air-cooled and liquid-cooled options depending on their deployment needs.
Alongside the H15 servers, Super Micro introduced new GPU servers equipped with AMD Instinct MI350P graphics processors, supporting up to ten GPUs in a 5U chassis without the need for upgraded cooling systems. This boosts AI training and inference performance while optimizing data center space and costs.
Networking enhancements come from AMD Pensando Pollara 400 AI network cards, which facilitate scale-out connectivity and maintain compatibility with standard Ethernet setups. This helps enterprises build larger AI clusters efficiently.
The company’s latest innovation, the Helios Platform, offers rack-scale AI computing power with up to 72 GPUs, designed to handle massive workloads for hyperscale operators and cloud providers.
Following the announcement, SMCI shares rose 1.54%, closing at $31.03, reflecting investor confidence in the expanded AI infrastructure strategy.
