, a provider of supercomputing solutions, has unveiled a new solution that combines the features of NVIDIA’s Tesla based 1U GPU system with that of an Intel (News
) Xeon based Appro server.
The Appro HyperPower Cluster offers a scalable computing architecture that can execute thousands of concurrent throughput parallel processing threads for mathematically intensive problems.
According to Appro, the solution is built from the company’s own high density servers and NVIDIA (News
) Tesla S1070 servers to include interconnect switches for node-to-node communication, master node, and clustering software all integrated in a 42U standard rack configuration.
The Appro HyperPower clusters start with a 1U chassis with two so-called "twin servers" inside. These are two-socket, half-width servers that support Intel's current "Nehalem EP" Xeon 5500 processors. Each Nehalem server is linked to a server appliance that has a single Tesla S1070 GPU in it. The Tesla S1070 packs four GPUs - each with 240 cores running at between 1.3 GHz and 1.44 GHz - into a single server chassis that also has 16 GB of its own memory.
The Appro HyperPower puts 19 of the twin Nehalem EP servers interleaved in a standard 42U rack with 19 of the Tesla appliances, which yields 304 x64 cores and 18,240 GPU cores. The peak performance of such a rack weighs in at just over 78 teraflops on single-precision codes and 6.56 teraflops at double-precision math.
While many companies have made attempts to cash in from the Tesla GPUs and their related CUDA tools to develop their own versions of personal supercomputers. Appro’s VP of Advanced Technology Solutions (News
), John Lee has commented that Appro waited for the Tesla technology to mature. The latest Tesla cards now offered by NVIDIA are integrated with double precision math which is very essential for workloads.
Lee says that customers want to start deploying hybrid x64-GPU systems using the CUDA environment today so they will be ready for the next generation of nVidia GPU co-processors.
Appro’s new clusters are capable of addressing the growing demand among customers for new GPU hybrid processing solutions that combine the functions of standard CPUs with GPUs, accelerators and FPGAs (field programmable gate arrays) and can execute thousands of concurrent throughput parallel processing threads, pushing science and research projects to the next level of computing.
Nathesh is a contributing editor for TMCnet. To read more of Nathesh's articles, please visit his columnist page.
"We are excited to work with Appro to broaden our high performance computing customer base through the provision of easy to deploy, ready to power up clusters," said Andy Keane (News
), general manager of the Tesla Business Unit at NVIDIA. "Together, we look forward to extending the reach of NVIDIA Tesla technology into a wide range of HPC customers who can now benefit from unparalleled supercomputing performance at lower price points."
Edited by Tim Gray