Super Micro Computer, Inc. has announced its next-generation H15 server portfolio powered by 6th Gen AMD EPYC 9006 Series CPUs. Designed to support advanced GPUs including AMD Instinct and connected via AMD Pensando networking, the updated systems deliver up to 256 cores and 512 threads to handle the rapidly growing demands of cloud, enterprise, storage, high-performance computing, and agentic AI workloads.

The updated lineup offers a notable 1.7x generational CPU performance improvement based on SPECInt Rate 2017 results published by AMD. By expanding memory and I/O bandwidth while increasing compute density, these systems allow organizations to run more concurrent AI agents and accelerate applications within existing power envelopes. The new platforms feature 33 percent more cores, twice the PCIe bandwidth, and 2.6 times higher memory bandwidth compared to previous generations.

“The latest AMD powered additions to our DCBBS family deliver the next generation of AI infrastructure, optimized for high performance, rapid scalability, and peak efficiency,” said Vik Malyala, Chief Business Officer at Supermicro. “Backed by our global services team, resilient U.S. supply chain, and consistent investment in US AI innovation, we continue to help customers deploy and scale AI with confidence.”

The new H15 portfolio includes tailored server architectures for varied enterprise needs. The flagship dual-socket Hyper platform is engineered for enterprise applications, AI inference, and cloud workloads. CloudDC targets cloud-scale environments using open standards, while the GrandTwin and FlexTwin high-density architectures provide power-efficient multinode compute for hyperscale deployments. Supermicro has also included Petascale Storage options supporting up to 4.8 PB per system for AI data lakes, alongside the 8U SuperBlade platform available in air-cooled and liquid-cooled configurations.

“As enterprises scale agentic AI, they need infrastructure that delivers exceptional performance, efficiency, and flexibility,” said Dan McNamara, senior vice president and general manager, Compute and Enterprise AI, AMD. “By combining the latest AMD EPYC CPUs, Instinct GPUs, and AMD Pensando networking with Supermicro’s modular server and rack-scale designs, customers can deploy AI infrastructure faster while improving utilization, reducing energy consumption, and lowering total cost of ownership.”

In addition to the H15 server lineup, Supermicro is expanding its AMD GPU-powered AI infrastructure. The new 5U AS -5126GS-TNRT and AS -5126GS-TNRT2 air-cooled systems support up to ten AMD Instinct MI350P PCIe GPUs featuring up to 144GB of HBM3e memory. Network integration is further enhanced by the AMD Pensando Pollara 400 AI NIC, which provides high-performance, open Ethernet scale-out connectivity.

For massive deployments, Supermicro and AMD are collaborating on the Supermicro AMD Helios Platform. This liquid-cooled, 72-GPU rack-scale platform incorporates AMD Instinct MI455X GPUs, 6th Gen AMD EPYC processors, AMD Pensando networking, and the AMD ROCm software stack. Built on Data Center Building Block Solutions architecture, these validated systems enable enterprises to deploy scalable AI infrastructure from single units to full-scale data center environments.