Infrastructure Management for High Performance Computing: Why Bright Leads the Way

    

lee-carterWith over 400 customers using Bright technology to manage their high performance computing environments, there’s no doubt that organizations around the world turn to Bright to solve a real business challenge.

We started out in 2009, and it’s easy to understand why early adopters of Bright technology were educational and research establishments. With a requirement for huge computational power, the benefit of Bright’s ability to install a complete cluster from bare-metal to turnkey supercomputer in as little as an hour was clear. I don’t know anyone - or any company - that can do it faster. And with Bright, these research-intensive organizations are able to provision, operate, monitor, manage, and scale clusters of any size.

However, Bright doesn’t stop at research houses; we serve many industry verticals and organizations of all sizes. Our 400+ customers represent multi-nationals, manufacturing, aerospace, oil and gas, telecommunications, software giants, life and health sciences, government agencies, and more. All of these organizations have a fundamental goal: to simplify the management of their clusters – be they centralized or distributed – and focus their energy on their business or research goals.

So, let’s take a look at some of the most common reasons customers choose Bright:

Bright empowers infrastructure consolidation

A major manufacturing company approached Bright with a requirement to consolidate its IT environment. With 13,000 servers distributed amongst multiple departments spanning multiple geographies, administration was ad hoc, maintenance insufficient, and there was a constant risk of being exposed to security risks.

Bright was chosen to provision, monitor, manage, and report on all aspects of servers including power and environmentals, ancillary hardware components, the Linux operating environment, and the various engineering application stacks, and seamlessly integrated with the company’s existing standards for security, storage, and workload management.

As a result, the company reduced its server count from 13,000 to 6,500, and utilization tripled to 90% (on average). Read more.

Bright scales to support clusters of any size

In January this year, we announced that the world's most powerful supercomputer dedicated to life sciences research had chosen Bright infrastructure management technology. The supercomputer - known as Computerome - is owned by CBS (Center for Biological Sequence Analysis), and ranks 121 in the list of the world’s largest publicly known supercomputers.

With more than 16,000 cores, 92 TeraByte of memory, and 7.5 PetaByte of storage, great emphasis is placed on using innovative technologies to optimise the power of Computerome. Bright provides the overall system software and management solution that powers Computerome, and was chosen for its flexibility and remote management capabilities.

As a result, the systems management team is equipped with a robust, responsive and comprehensive single-pane-of-glass management interface for Computerome’s hardware, operating system, HPC software, and users. Read more

Bright makes hybrid computing easy

Virginia Bioinformatics Institute (VBI) provides informatics and integrated analytics to advance human health, well-being, and security. VBI needed a powerful compute engine capable of handling very large datasets in memory, using hardware that met the particular demands of large-scale informatics research.

The institute built a powerful hybrid compute cluster comprised of Dell™ PowerEdge™ R and C series servers, NVIDIA Tesla™ GPGPUs, and Convey FPGA systems. Mellanox and Intel/QLogic InfiniBand switches provide high-speed connections between the servers and the storage. VBI built and tested the cluster, using Bright Cluster Manager in less than five days. Now that it’s operational, VBI uses Bright to provision, schedule, monitor, and manage the clustered environment for HPC. Bright’s single solution and integrated approach means that only two full-time employees are needed to manage this hybrid-architecture resource.

Bright enables VBI to manage this large, complex cluster serving diverse requirements without bringing on additional headcount. With Bright, this complex cluster is managed through an easy to use management interface, allowing end users to focus on other priorities and projects. Read more

###

We are living in an age where the pace of technological advancement is unprecedented. As a result, the requirement for (and reliance upon) high performance computing is becoming more critical than ever before. Bright is working alongside its customers, to support their innovation, speed their research cycles, power their computations, and fuel their growth.

High Performance Computing eBook