Supermicro Partners with Arm to Launch Energy-Efficient AI Infrastructure Solutions
In the ever-evolving landscape of artificial intelligence, Super Micro Computer, Inc. (NASDAQ: SMCI) is making a significant stride by partnering with Arm® to unveil a new category of energy-efficient rack-scale infrastructure solutions aimed at enterprise Agentic AI applications. This advancement comes as a direct response to the escalating computational requirements posed by modern agentic AI, necessitating a fresh approach to data center infrastructure that maximizes performance while minimizing energy use and physical footprint.
Supermicro's latest offerings leverage Arm's AGI CPU technology, specifically designed for orchestrating complex AI workloads effectively. These innovative solutions are engineered to support the rapid growth of agentic AI, delivering enhanced performance and efficiency that maximize rack usage in enterprise data centers. Moreover, by integrating Supermicro's Data Center Building Block Solutions® (DCBBS), organizations can expect a reduced time-to-online (TTO) for large-scale AI infrastructure deployments.
According to Charles Liang, President and CEO of Supermicro, "Supermicro continues to lead the industry in deploying innovative rack-scale solutions that maximize performance and efficiency. Our DCBBS technology stack provides a comprehensive range of data center solutions, and when paired with the high-density, efficient performance of the optimized Arm AGI CPU microarchitecture, organizations investing in agentic AI can greatly reduce their total cost of ownership (TCO)."
As identified by Mohamed Awad, Executive Vice President of Arm’s Cloud AI Business Unit, agentic AI represents a fundamental shift in infrastructure requirements, emphasizing efficiency, scalability, and orchestration performance in addition to raw computing power. The Supermicro and Arm collaboration aims to deliver infrastructure solutions tailored for achieving higher throughput in AI tasks while ensuring optimal data center economics at scale.
Supermicro’s new computer platform includes air-cooled 2U rackmount servers optimized for computing power and 5U rackmount servers tailored for GPU tasks. These platforms provide flexible designs that accommodate various applications and workloads. One specific innovation is a liquid-cooled multi-node solution crafted for agentic AI implementations at rack scale. This integrated approach takes advantage of Supermicro's proven modular, high-density architectures coupled with Arm's energy-efficient Neoverse® CSS V3-based CPUs.
Estimates suggest that the Arm AGI CPU deployed in Supermicro solutions can offer over double the performance per rack when compared to traditional architectures. This means significant savings, potentially amounting to $10 billion in capital expenditures (CAPEX) per gigawatt of AI data center capacity. These advanced solutions build upon Supermicro's strong reputation for rack density and performance per watt, allowing customers to optimize their data center space and energy resources effectively.
The Arm AGI CPU features a compact microarchitecture packed with 136 cores, engineered for high performance, minimizing legacy system overhead while maximizing workload processing efficiency. With a memory bandwidth of 6 GB/s per core and optimized memory access for low latency, this architecture supports linear scalability benefits, providing high memory capacity and flexible I/O for a scalable, energy-efficient agent-based AI infrastructure capable of managing thousands of tasks concurrently.
Supermicro’s Arm-based server lineup includes five distinct models:
1. 2U Hyper Server: Optimized for AI and memory-intensive workloads.
- Features: Two Arm AGI CPUs (up to 136 cores each), up to 6 TB DDR5-8800 MT/s RDIMMs, up to two GPUs.
2. 5U GPU Server: Built for GPU-intensive AI training and inference.
- Features: Two Arm AGI CPUs, up to 136 cores each, up to 8 dual-width GPUs.
3. 2U4N Liquid-Cooled Server: Designed for OCP ORV3 environments.
- Features: Two Arm AGI CPUs per node (potentially 20,672 cores in a single ORV3 rack).
4. 2U Hyper-E Server: A single-socket architecture optimized for edge computing.
- Features: Single Arm AGI CPU, up to 136 cores, up to 3 TB DDR5-8800 MT/s RDIMMs.
5. 1U 4N in an OCP ORW rack: High-density computing.
- Features: 336 Arm AGI CPUs per rack, supporting 168 servers and 45,696 cores.
Supermicro is committed to maintaining its leadership in the industry with a comprehensive portfolio of AI infrastructure solutions, enabling scalable, efficient, and environmentally responsible implementations across data centers worldwide. The latest rack-scale solutions will be displayed at the Supermicro booth in Taipei Nangang Exhibition Center Hall 1, where attendees will have a chance to explore the design and capabilities of these innovative products.
Founded and headquartered in San Jose, California, Supermicro is a globally recognized provider of comprehensive IT solutions optimized for various applications. The company is focused on delivering first-to-market innovations in enterprise, cloud, AI, and 5G Telco/Edge IT infrastructure. Proudly offering a Total IT Solutions framework that includes servers, AI, storage, IoT, and supporting services, Supermicro continuously innovates to meet the demands of its global clientele while enhancing the sustainability of its operations through green computing practices.
Supermicro's latest offerings leverage Arm's AGI CPU technology, specifically designed for orchestrating complex AI workloads effectively. These innovative solutions are engineered to support the rapid growth of agentic AI, delivering enhanced performance and efficiency that maximize rack usage in enterprise data centers. Moreover, by integrating Supermicro's Data Center Building Block Solutions® (DCBBS), organizations can expect a reduced time-to-online (TTO) for large-scale AI infrastructure deployments.
According to Charles Liang, President and CEO of Supermicro, "Supermicro continues to lead the industry in deploying innovative rack-scale solutions that maximize performance and efficiency. Our DCBBS technology stack provides a comprehensive range of data center solutions, and when paired with the high-density, efficient performance of the optimized Arm AGI CPU microarchitecture, organizations investing in agentic AI can greatly reduce their total cost of ownership (TCO)."
As identified by Mohamed Awad, Executive Vice President of Arm’s Cloud AI Business Unit, agentic AI represents a fundamental shift in infrastructure requirements, emphasizing efficiency, scalability, and orchestration performance in addition to raw computing power. The Supermicro and Arm collaboration aims to deliver infrastructure solutions tailored for achieving higher throughput in AI tasks while ensuring optimal data center economics at scale.
Product Offerings
Supermicro’s new computer platform includes air-cooled 2U rackmount servers optimized for computing power and 5U rackmount servers tailored for GPU tasks. These platforms provide flexible designs that accommodate various applications and workloads. One specific innovation is a liquid-cooled multi-node solution crafted for agentic AI implementations at rack scale. This integrated approach takes advantage of Supermicro's proven modular, high-density architectures coupled with Arm's energy-efficient Neoverse® CSS V3-based CPUs.
Estimates suggest that the Arm AGI CPU deployed in Supermicro solutions can offer over double the performance per rack when compared to traditional architectures. This means significant savings, potentially amounting to $10 billion in capital expenditures (CAPEX) per gigawatt of AI data center capacity. These advanced solutions build upon Supermicro's strong reputation for rack density and performance per watt, allowing customers to optimize their data center space and energy resources effectively.
The Arm AGI CPU features a compact microarchitecture packed with 136 cores, engineered for high performance, minimizing legacy system overhead while maximizing workload processing efficiency. With a memory bandwidth of 6 GB/s per core and optimized memory access for low latency, this architecture supports linear scalability benefits, providing high memory capacity and flexible I/O for a scalable, energy-efficient agent-based AI infrastructure capable of managing thousands of tasks concurrently.
Supermicro’s Arm-based server lineup includes five distinct models:
1. 2U Hyper Server: Optimized for AI and memory-intensive workloads.
- Features: Two Arm AGI CPUs (up to 136 cores each), up to 6 TB DDR5-8800 MT/s RDIMMs, up to two GPUs.
2. 5U GPU Server: Built for GPU-intensive AI training and inference.
- Features: Two Arm AGI CPUs, up to 136 cores each, up to 8 dual-width GPUs.
3. 2U4N Liquid-Cooled Server: Designed for OCP ORV3 environments.
- Features: Two Arm AGI CPUs per node (potentially 20,672 cores in a single ORV3 rack).
4. 2U Hyper-E Server: A single-socket architecture optimized for edge computing.
- Features: Single Arm AGI CPU, up to 136 cores, up to 3 TB DDR5-8800 MT/s RDIMMs.
5. 1U 4N in an OCP ORW rack: High-density computing.
- Features: 336 Arm AGI CPUs per rack, supporting 168 servers and 45,696 cores.
Supermicro is committed to maintaining its leadership in the industry with a comprehensive portfolio of AI infrastructure solutions, enabling scalable, efficient, and environmentally responsible implementations across data centers worldwide. The latest rack-scale solutions will be displayed at the Supermicro booth in Taipei Nangang Exhibition Center Hall 1, where attendees will have a chance to explore the design and capabilities of these innovative products.
Company Background
Founded and headquartered in San Jose, California, Supermicro is a globally recognized provider of comprehensive IT solutions optimized for various applications. The company is focused on delivering first-to-market innovations in enterprise, cloud, AI, and 5G Telco/Edge IT infrastructure. Proudly offering a Total IT Solutions framework that includes servers, AI, storage, IoT, and supporting services, Supermicro continuously innovates to meet the demands of its global clientele while enhancing the sustainability of its operations through green computing practices.
Topics Business Technology）
【About Using Articles】
You can freely use the title and article content by linking to the page where the article is posted.
※ Images cannot be used.
【About Links】
Links are free to use.