Even with the rapid advancement of A.I. and deep learning technology, many enterprises still need help implementing these technologies to critical applications. Why? This is because it takes quite a bit of hardware to meet the computing demand for A.I. and deep learning workloads. In addition, specific resources are required to set up the training and analysis environments, and the training process can take a lot of work. This is where Supermicro’s new deep-learning solution for data centers comes into play.
Supermicro provides an end-to-end solution for enterprises to address, train, test, and deploy A.I./deep learning applications across multiple industries quickly and efficiently. Partnering with Canonical, OS4 Open Source, and Intel, Supermicro’s solution provides an easy-to-deploy, optimized and pre-configured hardware and software stack for deep learning. The solution significantly reduces TCO for A.I./deep learning ecosystem. The architecture can be deployed on-premise or in public cloud environments such as AWS and Azure.
The Supermicro deep learning solution complements the company’s overall portfolio of data center products. These include Supermicro® AI and Deep Learning Motherboards (Xeon Phi and NVIDIA GPU families), Supermicro SuperServer systems with optimized CPU and GPUs for deep learning, and SuperMicro solutions with Intel® FPGAs to accelerate applications such as training neural networks. Supermicro is strengthening its leadership in the data center market by taking advantage of A.I. and deep learning technology. Supermicro SuperServer motherboards are used to build a deep learning/A.I. cluster. That said, we should not conclude from the fact that this motherboard will be based on them and that it will support deep learning workloads.
Reliability and security are paramount for the deep learning solution. Supermicro has relied on high-performance Intel® Xeon Phi processor Xeons paired with NVIDIA Tesla and Quadro solutions to provide the required computing power. Supermicro SuperServers are optimized for deep learning and A.I. workloads. A customized version of Supermicro’s high-performance server solution is also integrated with NVIDIA® Tesla and Quadro motherboards to make it more scalable. This helps to increase the computing power for each node.
Supermicro SuperStorage systems provide enterprise storage systems that offer high TCO and data center deployment performance. Supermicro SuperSwitches give the flexibility and interpretation essential for data center workloads, which are enabled with support for QSFP28, 24G CXP, CX4, and 8G EDR links. As a public cloud provider, AWS is used with the Supermicro deep learning solution. The Intel® Arria® 10 field programmable gate arrays (FPGAs) are used in the solution that is optimized for deep learning and other data center workloads. The Canonical Distribution of Kubernetes provides a Linux container-based application platform that supports agility, portability, and composability across multi-cloud and hybrid deployments. It provides Kubernetes, container management, monitoring, and operations for enterprise customers across industries at scale.
With the growth of deep learning, Supermicro has developed its deep learning solution. Since the market is evolving so fast, the company had to do something to catch up with all the changes. These changes are happening not only in CPU and GPU solutions but also in software solutions used for data center deployment. Therefore, Supermicro has tailored software solutions for its mission-critical server systems for data centers.
The Supermicro deep learning appliance provides an easy-to-deploy, and scalable compute infrastructure across multiple industries. Supermicro also provides a turnkey, pre-configured solution with optimized hardware and software. Supermicro leverages its global presence, sales force, engineering, and manufacturing to ensure enterprises can deploy deep learning solutions sooner rather than later. This, combined with the integrated solution, reduces TCO for the customer.
The rich ecosystem of the Supermicro deep learning solution has allowed over 10,000 HPC systems to be deployed around the globe, with one of the most extensive installed systems bases worldwide. The Supermicro deep learning appliance is designed to run an A.I./deep learning workload on CPU, GPU, and FPGA. The solution can be deployed on-premises or in the public cloud.