SCIENCE

Boston Limited has formally announced the launch of their next generation of blade server solutions, the groundbreaking new Boston SuperFlex Blade.

Designed in partnership with Supermicro, the SuperFlex Blade brings to the industry a new dimension of energy efficient hybrid parallel compute processing power for technical and enterprise computing. SuperFlex Blade platforms have been optimised to offer a new dimension in flexibility for Blade Servers by providing 2 x PCI-Express x 16 expansion slots per blade with support for dual full-length and double-height expansion cards.

"Building off the success of the 1U GPU Superserver, the new SuperFlex Blade delivers high density GPU computing in a highly flexible and scalable solution," said Sumit Gupta, Manager of Tesla products at NVIDIA. "Through our collaboration with Boston and Supermicro, this solution takes full advantage of the massive computational power of Tesla GPUs and provides exceptional performance across a wide range of demanding applications."

Optimised for hybrid computing, the SuperFlex Blade is suitable for the most demanding applications requiring high-speed and high-density processing with maximized GPU and CPU compute performance. High performance networking has also been optimised with two independent, high-bandwidth, low-latency onboard QDR InfiniBand controllers standard to ensure that this platform is suitable for today's most resource-hungry computing challenges.

"The Boston SuperFlex Blade has been jointly developed with Supermicro to solve the key scalability and usability challenges that face customers in high performance computing (HPC) environments. Using the standard Supermicro 10-bay blade enclosure, it offers unparalleled compute density when equipped with two double-width GPU accelerators," says Manoj Nayee, Managing Director of Boston Limited.

Designed to meet the most demanding applications computing challenges, the SuperFlex Blade provides phenomenal levels of hybrid computational power by combining the latest technologies made available by Intel Xeon 5600 processors in conjunction with the latest generation of NVIDIA Tesla 2050 and 2070 GPU compute processors for maximum computational performance.

"This new GPU-optimized blade design offers a much higher level of computing performance, density and performance-per-watt," said Wally Liaw, Vice President of Sales-International at Supermicro. "By supporting up to twenty GPU cards in one 7U blade enclosure, this solution easily delivers the highest GPU density available today in a blade form factor."