“The OSS GPUltima is the densest and most cost-effective petaflop solution for scalable data center infrastructures,” said Steve Cooper, OSS CEO. “Supporting up to 128 NVIDIA Tesla K80 GPU accelerators, the OSS GPUltima shows tremendous performance gains in many applications like oil and gas exploration, financial calculations, and medical devices. Providing ample cooling and power to accommodate this many high-end cards, it surpasses other devices in performance. In addition, the OSS GPUltima is a preconfigured rack of GPUs, servers and interconnections already integrated and the whole unit is tested and ready to add application software.”
At SC15, One Stop Systems introduced the GPUltima petaflop compute platform. As a highly dense platform, GPUltima delivers petaflop performance at greatly reduced cost and space requirements. The OSS GPUltima supports up to 128 double-wide interconnected accelerators. Consuming only 56 kilowatts of power, the platform uses 94 percent less power than other petaflop solutions. This reflects a tremendous cost savings in power and cooling. In addition the single-rack occupies 96 percent less space than other solutions. The OSS GPUltima employs 64 OSS PCIe Gen3 x16 adapters and cables between the OSS servers and the OSS High Density Compute Accelerators (HDCA), each containing 16 high-performance NVIDIA Tesla K80 GPU accelerators and 32 Mellanox EDR 100Gbs InfiniBand adapters and cables connecting the GPUs through a 36-port InfiniBand switch. One Stop Systems’ PCIe expansion systems provide high speed connectivity between the CPU and IO cards. While IO cards have traditionally occupied slots in the server cabinet, OSS designs expansion appliances that support large numbers of high speed cards with latest technology PCIe Gen3 slots, efficient cooling, and sufficient power.
NVIDIA Tesla K80 GPU accelerators are the flagship offering of the Tesla Accelerated Computing Platform, the leading platform for accelerating data analytics, deep learning and scientific computing. It integrates the world’s fastest GPU accelerators, development tools, high-speed communication technology, a supported ecosystem and the NVIDIA CUDA parallel computing model. NVIDIA’s GPUDirect technology enables multiple GPUs to directly read and write CUDA host and device memory, eliminating unnecessary memory copies, dramatically lowering CPU overhead, and reducing latency, resulting in significant improvements in application performance.
Mellanox EDR 100Gb/s InfiniBand adapters, switches, cables and software are the most efficient solutions for server and storage connectivity, delivering high throughput, low latency and industry-leading application performance for enterprise solutions that are widely deployed across HPC, Cloud, Web 2.0, and Big Data satisfying the most demanding data centers requirements. The OSS GPUltima employs 32 EDR 100Gb/s InfiniBand adapters and a 36-port EDR 100Gb/s InfiniBand switch from Mellanox in each platform.
Mellanox has long been recognized as the leader in InfiniBand technology, RDMA and GPU offloading engines. Our solutions, providing the high-performance networking for the GPUltima, are an excellent testament to the performance, efficiency and application acceleration our latest end-to-end EDR 100Gbs InfiniBand technology provides,” said Gilad Shainer, vice president of marketing, Mellanox Technologies. “We are excited about being a part of the One Stop Systems team in developing the GPUltima.”
The OSS GPUltima equipped your choice of GPU accelerators is available immediately.