Technology newsroom
Supermicro launches new H15 server! The power of AMD EPYC 9006 Gen 6 is 1.7 times stronger! Fully prepared for AI and Enterprise 🚀
Supermicro launches new H15 server powered by 6th Gen AMD EPYC 9006 Series CPUs, delivering 1.7x higher performance with up to 256 cores/512 threads, specifically designed to support AI, HPC and Cloud applications.
📌 Summary of important points:
- Supermicro launches new H15 server powered by 6th generation AMD EPYC 9006 Series CPUs.
- Highlights CPU performance that is 1.7 times stronger compared to the previous generation. Comes with up to 256 cores and 512 threads.
- Designed to support next-generation AMD Instinct GPUs and AMD Pensando networking for AI, HPC, and Cloud applications.
Supermicro announces next generation H15 server Powered by the 6th Gen AMD EPYC 9006 Series CPU, designed to support next-generation AMD Instinct GPUs, and connected by dedicated AMD Pensando networking. With up to 256 cores and 512 threads, the H15 server can meet the exponentially increasing computing demands of Cloud, Enterprise, Storage, High-Performance Computing (HPC), and especially Agentic AI.
Highlights include a 1.7x increase in CPU performance compared to the previous generation, along with expanded memory and I/O bandwidth. Enables customers to run more AI agents simultaneously, accelerating enterprise applications and increase the efficiency of Host-Node to the maximum within the original energy limitations. This significantly reduces the total cost of ownership (TCO).
💻 Deep dive into the H15 Portfolio that answers every workload.
Supermicro has designed a variety of products in the H15 group to support different uses, such as
- Hyper: Dual-socket platform for Enterprise applications, AI inference, and Virtualization.
- CloudDC: Single/Dual socket server designed according to Open Compute Project (OCP) standards for cloud-scale environments.
- GrandTwin: High-density 2U 4-node architecture optimized for Object storage, Virtualization and Cloud services.
- FlexTwin: High-performance 10U 2-node system that uses Liquid Cooling to maximize power density and efficiency.
- Petascale Storage: High-capacity all-flash storage platform for AI data lakes and large-scale analytics
- SuperBlade: Blade architecture for HPC and AI inference that requires the highest density.
🚀 Expand your AI infrastructure with AMD GPUs
In addition to the CPUs, Supermicro is also announcing new 5U PCIe GPU servers (AS -5126GS-TNRT and AS -5126GS-TNRT2) that support up to 10 AMD Instinct MI350P PCIe GPUs with 144 GB of HBM3e memory to accelerate AI processing in both Inference and Training. Additionally, the Supermicro AMD Helios Platform is a rack-scale solution with AMD Instinct GPUs. Up to 72 MI455X units with liquid cooling Specifically for training large-scale AI models
🌐 High performance networking with AMD Pensando
To ensure seamless connectivity, Supermicro has deployed AMD Pensando Pollara 400 AI NIC, a high-performance Open Ethernet network card. To provide high bandwidth and low latency connectivity between servers and storage for AI Clusters, Supermicro's Data Center Building Block Solutions (DCBBS) offers a comprehensive and well-tested solution. Helps organizations deploy AI more quickly Save energy and reduce operating costs in the long run
💬 How do you think CPUs with this many Cores/Threads will change the face of the Data Center and AI industry in the future?