Introducing New Supermicro Solutions Featuring Arm AGI CPUs
Optimized architectures unlocking breakthrough performance, efficiency, and rack density for 24/7 agentic AI
Supermicro solutions featuring Arm AGI CPUs are built for the age of massive-scale agentic AI orchestration, with core performance, efficiency, and density that maximizes the economics of rack-scale deployments. Supermicro’s proven first-to-market leadership combines with large-scale DCBBS deployment capabilities and in-house developed thermal management technologies to create a new class of architecture optimized for modern agentic AI. Leveraging the Arm AGI CPU’s AI-centric, memory-optimized microarchitecture, Supermicro solutions deliver exceptional, scalable performance that efficiently orchestrates agentic AI workloads at scale.
Deploy AI-Scale Infrastructure Faster with Supermicro
The Supermicro first-to-market advantage backed by modular, workload-specific architectures, liquid cooling & rack integration expertise, and a global manufacturing footprint
Supermicro DCBBS for Complete Agentic AI Solutions
The blueprint for modern data centers, Supermicro DCBBS provides complete, modular infrastructure built from validated components and subsystems
Compute for a New Age of Agentic AI
All-new purpose-built Arm AGI microarchitecture delivers up to 2x performance and density compared to x86 for significant power and space savings at rack scale
Supermicro Rack-Scale Agentic AI Solutions
Combining Supermicro’s proven modular, high-density architectures with the energy-efficient Arm Neoverse CSS V3-based CPUs enables scalable, flexible infrastructure that maximizes performance-per-watt and dramatically lowers energy demand to accelerate AI adoption across modern data centers. By leveraging Arm AGI CPU, Supermicro solutions can deliver over 2x performance per rack compared to traditional microarchitectures and help enterprises save up to $10 billion in CAPEX per Gigawatt of AI data center capacity. Building on Supermicro's industry-leading rack density and performance-per-watt, these solutions help ensure maximum utilization of data center space and power resources.
| Product(s) | 2U 4-Node Rack-Scale Solution Ultra-density multi-node rack scale liquid cooled ORV3 rack architecture |  1U 4-Node Rack-Scale Solution Maximum compute density multi-node liquid cooled ORW rack architecture |
|---|---|---|
| Key Applications | Large-scale AI Inference, Cloud Compute, Hyperscale | Large-scale AI inference, Agentic AI Orchestration, Cloud Compute, Hyperscale CPU Services, High-density Scale-out Workloads |
| Rack Configuration |
|
|
| CPUs/Cores per Rack | Up to 152 CPUs/20,672 cores | Up to 336 CPUs/45,696 cores |
| Cooling Solution |
|
|
| Power Budget per Rack | Max 95kW | Max 210kW |
| Webpage Link | Learn More | Coming Soon |
Supermicro Rackmount Systems
Supermicro leads the industry in bringing new and transformative technologies to market, leveraging a modular building block philosophy and rich expertise in design and development, from the system level right up to data center scale. These new and innovative, workload-specific solutions are based on proven architectures designed to maximize the performance and density of Arm AGI CPUs while also ensuring power and thermal efficiency, lowering TCO and TCE (total cost to the environment).
| Product(s) |  5U GPU Server Thermally-optimized Architecture with Flexible GPU Support |  2U Hyper Flagship Rackmount for a New Generation of Performance and Efficiency |  2U Hyper-E Single-socket Architecture Optimized for Edge Deployments |
|---|---|---|---|
| Key Applications | GPU-accelerated Inference, AI Training, Model Development, HPC, Simulation, Rendering | AI Orchestration, CPU-driven Inference, Memory-intensive Workloads, Cloud Compute, Virtualization, Flexible GPU-accelerated Inference | AI Inference, Edge Computing, Cloud, CDN |
| CPU | Dual Socket Arm AGI CPU Neoverse V3 with 64/128/136 cores | Dual Socket Arm AGI CPU Neoverse V3 with 64/128/136 cores | Dual Socket Arm AGI CPU Neoverse V3 with 64/128/136 cores |
| GPU Support | Up to 8 double-width | Up to 2 double-width | Up to 2 double-width |
| Storage | 8x 2.5" NVMe | 8x 2.5" NVMe | 6x E1.S or 2.5" NVMe |
| Webpage Link | Learn More | Learn More | Learn More |
Featuring the World’s Most Efficient Agentic CPU
The Arm AGI CPU is the first production silicon from Arm, designed for AI infrastructure at scale. It is designed to deliver a new class of CPU with high performance and extreme rack-level density support for agentic AI operations across modern data centers. The design choices of the Arm AGI CPU are made to deliver maximum performance at rack scale. From microarchitecture to memory, clock frequency to I/O—everything adds up to more performance at gigawatt scale.
| Feature |  Arm AGI CPU |
|---|---|
| Core Count | Up to 136 Neoverse V3 |
| CPU Architecture | Armv9.2 |
| System-level Cache | 128MB |
| Max Frequency | 3.5-3.7GHz |
| Base TDP | 300W |
| Memory |
|
| PCIe |
|
Efficiency
Purpose-built, no legacy overhead No wasted energy or silicon
Scale
Linear scaling across cores Memory and I/O keep pace
Performance
More work per cycle Full performance, always