Designed for a Faster, Better, and Greener Hyperscale Infrastructure from the Edge to the Cloud
Today, the modern data center constantly struggles the balance between standardization for resource optimization and flexibility for business competitiveness. With the ever-growing demand of computing power, hyperscale data centers have faced challenges to deliver a highly versatile solution optimized for cost and performance while maintaining the flexibility and openness based on the commercial off the shelf (COTS) standard. Supermicro’s Optimized AI servers, along with its MegaDC and CloudDC product families, are the answer to the challenge. The Supermicro AI servers are designed for both air and liquid cooling and are based on OCP Inspired specifications. MegaDC and CloudDC are the key solutions for your next-generation hyperscale data center. With extended support of open standards, including OpenBMC for version control, AIOM modules that support the OCP3.0 SFF standard, and power-optimized design, data center operators can enjoy the benefit of an open compute concept without revamping their existing infrastructure.
The Open Compute Project Foundation (OCP) was initiated in 2011 with a mission to apply the benefits of open source and open collaboration to hardware and rapidly increase the pace of innovation in, near and around the data center’s networking equipment, general purpose and GPU servers, storage devices and appliances, and scalable rack designs. OCP’s collaboration model is being applied beyond the data center, helping to advance the telecom industry & EDGE infrastructure.
Supermicro is a Platinum Member of OCP and has advisory positions today. Supermicro continues to innovate on system design for data center deployments. Supermicro has a number of products which are “OCP Inspired”. These include a range of servers with Supermicro’s AIOM with its OCP 3.0 compliance. In addition, Supermicro leads the Industry in the number of systems that are part of the AI Marketplace.
Supermicro AI System Advantages
Supermicro designs and manufactures a wide range of systems, with many systems specifically designed for AI computing.
The new AI Marketplace Portal by OCP aspires to serve as the central hub for AI cluster designers and builders. Located in the OCP Marketplace, this portal consolidates a number of different technologies that are needed for a complete data center buildout.
Supermicro leads the industry in the design and delivery of AI Servers Based on OCP Inspired technologies.
Latest GPU Technology
With these servers from Supermicro, the latest GPU technology is available. Whether liquid cooled or air cooled, these servers are being used worldwide in some of the largest AI deployments on earth.
Improves Performance
The tight integration of GPUs with each other leads to significant performance improvement compared to PCIe-based hardware. For many applications, the GPU-to-GPU connection is critical to meet performance requirements.
Building Block Solutions®
Supermicro has the ability to create new servers in a variety of configurations, using its Building Block Solution® approach. The HGX and OAM-based GPU servers enable Supermicro and, thus, its customers to become more productive sooner.
By leveraging OCP 3.0 design concepts, Supermicro is implementing AIOM into its architecture to offer an of-the-shelf standard solution with improved architecture:
Thermal Improvement
Since AIOM cards are designed to be installed at the same level as the motherboard (see the following diagram), versus the traditional PCIe which is installed vertically or horizontally on top of the motherboard, AIOM significantly increases the airflow throughout the system, therefore improving thermal management.
Easy Serviceability
Because the AIOM card is installed from the rear of the chassis (see the following diagram), versus the traditional PCIe card that is installed from the top, the opening of the chassis top cover is no longer needed to install or remove the AIOM card. With the combination of the pull tab and screw knob, it eliminates the need for tools for service.
TCO Optimization
With its SFF (Small Form Factor), thermal efficiency, easy serviceability and implementation of OpenBMC, Supermicro is offering a solution that decreases time to service and minimizes system down time. By adapting a solution with AIOM from the server to the rack and to the data center level, as the infrastructure grows the improvement to TCO will grow together.
Discover How Supermicro OCP Solutions Reduce Data Center TCO and Direction for Supermicro
