SCIENCE

Bright Computing announced that GraphStream is now shipping Bright Cluster Manager as an integral part of its new generation of scalable platforms. The new GraphStream systems incorporate the latest NVIDIA Tesla and Quadro GPUs, and deliver unprecedented efficiency for both computation and visualization.

"Bright Cluster Manager is a key component of GraphStream's new platforms," said Teri Dunwoody, CEO of GraphStream. "From pre-shipment configuration and testing, through system commissioning and production operations at our customers' sites, GraphStream and Bright Computing are working together to transform the concept of turn-key HPC systems from a mere marketing claim into reality."

GraphStream partnered with NVIDIA in 2003 to deliver the world's first commercially-integrated GPU-based HPC clusters, and since then has deployed many of the world's most powerful systems of this type. GraphStream's superior depth of experience and in-house systems-engineering expertise specific to scalable GPU-accelerated computing, combined with global support capabilities up to 24x7x365 4-hour mission-critical on-site response now covering more than 50 countries, contributes to a compelling value proposition.

"Bright Cluster Manager was the clear choice for the new GraphStream platforms," added Dunwoody. "Compared with other cluster management software options, Bright Cluster Manager leaves more system resources available for applications, and the single per-node daemon enables efficient scaling to much larger node counts. Additionally, Bright's intuitive GUI and scriptable command-line interface provide system administrators with an unprecedented level of visibility and control."

"Bright Computing is pleased to work with GraphStream to deliver a new generation of HPC systems. These platforms combine the power of GPUs with the world's most advanced management software, to set a new standard for operational efficiency and ease of use," said Dr. Matthijs van Leeuwen, CEO of Bright Computing. "With a rapidly expanding base of applications that can take advantage of GPUs, it is now possible to speed application performance 2x, 100x or even faster -- truly revolutionary performance improvements, and at price points available to a much broader base of mainstream technical computing users."