In order to put more computing power in the hands of customers who need powerful supercomputers and Artificial Intelligence (AI) platforms, Lenovo has expanded its support for multiple graphics processing units, (GPUs), from Nvidia.   Select Lenovo systems will now be supporting the NVIDIA® Tesla™ M10, P40, and P100 GPU accelerators and NVIDIA deep learning platform for traditional HPC, as well as emerging AI deep learning applications.

More Power, More Answers

Tesla P100 for PCIe based servers are ideal for both HPC and deep learning training deployments. The P100 is powered by the powerful NVIDIA Pascal™ architecture, with 16GB of onboard memory, for demanding HPC applications.  With up to 4.7 TeraFLOPS of double-precision performance, a single P100 node can replace up to 32 traditional CPU nodes. 

Designed for inferencing, the Tesla P40, also powered by Pascal architecture, allows customers to use established patterns within neural networks to recognize items like images, text or speech.  Equipped with specialized inference instructions based on 8-bit integer (INT8) operations, the P40 delivers up to 22 TOPS (tera operations per second) of performance for inferencing use cases.

The Tesla M10 is designed to accelerate graphics in virtual desktop and application environments.  By utilizing NVIDIA GRID™ software, which enables the virtualization of physical GPUs into full-featured virtual GPUs, the M10 provides maximum performance and scalability of up to 64 users per board.

Beginning this month, all three GPUs will be supported on Lenovo’s x3650 M5, Lenovo’s flagship 2U rack server and on the nx360 M5, the ultra-dense compute node for our NeXtScale system that is at the heart of many of the world’s great research centers.

A Worldwide Impact

Currently, at 20 of the 25 top research universities worldwide, (according to the 2016 Elsevier Research), Lenovo systems are powering life-changing research across multiple fields of study.  For example:

At the University of Adelaide in Australia, Lenovo NeXtScale systems augmented with the NVIDIA GPU computing platform are helping to study particle physics and the theory of Quantum Chromodynamics (Quarks).  Just for balance, they are also studying global human happiness as reflected in the geographic use of Twitter®.   

At the University of Chicago, cancer researchers using Lenovo systems are working on using Magnetic Resonance (MR) images of prostate cancers to derive quantitative parameters which can correlate with cancer detection.  Their goal is to create risk maps of cancer within prostate to determine the optimal course of treatment

From quarks to Tweets to cancer research, Lenovo systems with the NVIDIA GPU computing platform power some of the most important research in the known universe.