AMD AAI 2026: Pegatron Accelerates Enterprise AI at Scale with 6th Gen AMD EPYC Server CPUs and AMD Pensando Pollara 400 AI NIC

Pegatron, a recognized provider of comprehensive AI infrastructure and server solutions, participated to the AMD Advancing AI 2026 conference, that took place July 22–23 at Moscone West in San Francisco, showcasing its latest portfolio of AMD advanced AI, HPC, and enterprise server platforms designed to accelerate the deployment of next-generation AI factories and modern data centers.Pegatron’s latest AMD – based portfolio combines the newly launched 6th Gen AMD EPYC Server processors, AMD Pensando Pollara 400 AI NIC, and AMD Instinct MI355X GPUs to provide a flexible, scalable foundation for enterprise AI, agentic AI, HPC, cloud, and data-intensive workloads.

“As AI deployments continue to scale, customers are looking for a broad range of AI infrastructure solutions,” said Gary Cheng, president and co-CEO of Pegatron. “Leveraging Pegatron’s expertise in rack-scale system design, system integration, and advanced cooling, together with AMD’s latest 6th Gen AMD EPYC processors, AMD Pensando Pollara AI NIC, and AI infrastructure technologies, we help customers build AI deployments with higher efficiency, scalability, and operational resilience.”

Next-Generation High-Density Computing with 6th Gen AMD EPYC Server processors
The Pegatron GS205-2A2 is a versatile 2U dual-socket platform for enterprise AI, HPC, virtualization, and data-intensive workloads. Advanced by the latest 6th Gen AMD EPYC Server processors, the system features an air-cooled thermal design rated for up to 2 x 600W CPU TDP, delivering up to 256 cores per socket — 33% more than the prior generation — along with high-speed DDR5 memory for maximum throughput and system performance. The platform delivers up to 1.87x higher performance versus the previous generation, along with an average 20% improvement in performance-per-dollar value generation-over-generation, driving greater efficiency and lower TCO.

Built on an industry-leading PCIe Gen 6 I/O subsystem offering 2x bandwidth per link versus the previous generation, the GS205-2A2 delivers up to 800Gb/s of total network bandwidth, providing ample headroom to support the AMD Pensando Pollara 400 AI NIC in an OCP 3.0 form factor alongside a PCIe-based NIC, or even next-generation network adapters, simultaneously for maximum networking flexibility. PCIe Gen 6 also accelerates storage, enabling PCIe Gen 6 SSDs with data throughput of up to 28GBps to keep pace with demanding AI and data-intensive workloads. Supporting up to two PCIe GPUs, the GS205-2A2 delivers the performance and flexibility needed for AI inference, visualization, agentic AI, cloud infrastructure, and modern enterprise applications.

“Enterprise AI has moved from experimentation to production, and agentic AI will place even greater demands on compute, memory bandwidth, efficiency and system-level orchestration,” said Dan McNamara, SVP and GM, compute and enterprise AI, AMD. “With 6thGen AMD EPYC Server CPUs, AMD is helping customers scale enterprise AI with the performance and efficiency needed to support increasingly complex data pipelines, AI services and intelligent applications across the modern data center.”

Accelerated AI and HPC with AMD Instinct MI355X GPUs
Pegatron is also showcasing the AS404-2A1-8I2, a 4U liquid-cooled AI server designed for large-scale AI training, inference, and HPC deployments. Advanced by dual AMD EPYC 9005 Series processors and eight AMD Instinct MI355X GPUs and also supporting the AMD Pensando Pollara 400 AI NIC for scale-out networking, the platform delivers exceptional AI performance with high memory capacity, flexible PCIe® Gen5 expansion, and advanced liquid cooling for maximum thermal and power efficiency.