Tom Coull, President & CEOThe data center landscape is witnessing an increased adoption of open technologies to make better use of a flexible, cost-effective scale-out architecture. These open technologies free organizations from the explicit costs and closed architectures of traditional vendors. Facebook’s Open Compute Project (OCP) is one such initiative that provides a collaborative development model for a wide range of companies to continue the path of innovation for storage, server, and data center. An active member of the OCP community, Penguin Computing provides built-to-order data center solutions based on the Open Rack standard and OCP principles. “All of our OCP solutions are engineered for maximum power efficiency, serviceability, and cost effectiveness from the ground up,” explains Tom Coull, President and CEO, Penguin Computing.
The company’s prime solutions fulfill four basic needs of HPC and data centers— open computing, storage, networking, and cloud solutions. Based on the open computing approach is the company’s Tundra Extreme Scale (ES) Series of servers that enables clients to integrate a wide variety of architectures in a single solution. Penguin Computing’s engineers can configure and customize the client’s environment with a variety of operating systems and management solutions including both open source and commercial applications. Likewise, support for multiple open software storage technologies including Lustre, Ceph, Gluster and Swift can be delivered through Penguin’s high-performance FrostByte Storage. FrostByte proactively monitors system health and in the event of a fault, immediately notifies the system administrator and Penguin Support.
Among the company’s core offerings are its Arctica products—enterprise-class software-defined networking solutions that enable customers to meet large-scale and complex data center Dev-Ops requirements. Through the Arctica switches, clients get tools to easily build, manage, and maintain large scale network deployments with high levels of programmability and choice of workflow.
In addition, Penguin’s open cloud-based solutions offer clients the ability to easily build and manage private clouds for HPC. Penguin’s Public Cloud, POD empowers clients to perform computational jobs easily and monitor them from either a traditional Linux CLI environment, an intuitive and secure web portal or a state-of-the-art, fully interactive remote workstation called Scyld Cloud Workstation.
All of our OCP solutions are engineered for maximum power efficiency, serviceability and cost effectiveness from the ground up
POD is ready-to-run with numerous applications pre-installed eliminating the complexity associated with building, managing, and scaling HPC environments. “We complement our hardware and software solutions with POD, providing supercomputing capabilities on-demand on a pay-as-you-go basis,” asserts Coull.
Alongside the four solutions, the company also provides professional and managed services to optimize and simplify computing operations. Penguin’s products and services together have benefitted a number of clients from industries including finance, government, manufacturing, Web content and education. In one example, the U.S. Department of Energy (DoE) required compute capabilities to support the National Nuclear Security Administration. Penguin Computing was awarded the contract to install Tundra ES series for high-performance computing clusters to ensure the safety, security, and reliability of DoE’s project. “Through Tundra ES system, we were able to combine the benefits of Open Compute with high-density compute architecture to meet the demanding supercomputing needs of an advanced program,” says Coull.
Moving ahead, the company will vest its focus on ushering OCP efficiency and innovation into the HPC, cloud and data center scale-out worlds through product enhancements and focused customer service. “We are also heavily investing in our public cloud to offer more capacity and capabilities–this is a key component in our roadmap,” concludes Coull.