Power8 Systems with NVLink Come to Nimbix HPC Cloud

Print Friendly, PDF & Email

nimbixToday Nimbix announced plans to expand its performance leadership in cloud supercomputing with the addition of IBM’s recently-launched Power Systems S822LC for HPC platform featuring NVIDIA Pascal GPUs. Nimbix will now put the capabilities of IBM Power Systems for HPC, based on the POWER8 with NVIDIA NVLink interconnect technology processor, and Nimbix’s JARVICE platform in the hands of developers and data scientists to achieve unprecedented performance.

Today’s emerging workloads like machine and deep learning, artificial intelligence, accelerated databases, and high performance data analytics require incredible speed through accelerated computing,” said Sumit Gupta, Vice President, High Performance Computing and Data Analytics, IBM. “Delivering the capabilities of the new IBM POWER8 with NVIDIA NVLink-based system through the Nimbix cloud expands the horizons of HPC and brings a highly differentiated accelerated computing platform to a whole new set of users.”

The new IBM Power Systems S822LC for HPC stems from community innovation within the OpenPOWER Foundation and is fine-tuned for deep learning, artificial intelligence and cognitive computing. The new POWER8 with NVIDIA NVLink processor directly connects the POWER8 CPU with NVIDIA Tesla P100 Pascal GPUs through NVIDIA NVLink, a high-speed, energy-efficient bidirectional interconnect. NVIDIA NVLink is embedded at the silicon level and incorporated into the overall system design. With this tight coupling of IBM and NVIDIA technology data can flow 5x faster than on a comparable competitive offering.

Offering this new platform on the Nimbix cloud enables the highest performance and ease of use for researchers, data scientists, and engineers looking to leverage optimized workflows. With a few mouse clicks, users can run massive computations such as accelerated analytics and neural network training without having to understand and configure toolkits and infrastructure first. Sample ready-to-run workflows include GPU-accelerated database systems and deep learning frameworks for both training and inference.

Additionally, developers and application providers can leverage Nimbix PushToCompute technology to rapidly build and deploy new hyper-scale workflows and frameworks targeting NVIDIA Pascal GPUs at bare-metal speed. Enterprises can also seamlessly extend existing accelerated applications into a hybrid cloud model combining powerful on-premises systems with on-demand cloud.

In the new era of accelerated analytics and deep learning, delivering both applications and bare-metal performance in the cloud with state of the art infrastructure has never been more important,” said Nimbix CEO Steve Hebert. “We’re pleased to partner with IBM and NVIDIA to democratize access to POWER, NVLink in combination with Nimbix PushtoCompute.”

“HPC scientists and AI researchers are among the most demanding users of computing resources,” said Roy Kim, Tesla Product Lead at NVIDIA. “Powered by Pascal-based NVIDIA Tesla P100 GPU accelerators and NVLink, the Nimbix cloud platform offers scientists and researchers on-demand, scalable compute resources to deliver significantly faster results than was previously thought possible.”

Nimbix provides on-demand and scalable compute resources that enable organizations to run large-scale HPC workloads in the cloud. HPC allows scientists, developers, and engineers to solve complex science, engineering, development, big data, and business problems using applications that require high compute capabilities. Nimbix’s growing number of integrated platform services allows organizations to increase the speed and effectiveness of research by running high performance computing in the Nimbix Cloud.

Sign up for our insideHPC Newsletter