With a substantial number of high-performance computing (HPC) applications combined with GPU acceleration, Nvidia releases new software and tools on the Nvidia GPU Cloud (NGC) platform, enabling scientists to deploy computing applications and HPC visualization tools quickly.
Nvidia Adds Cutting-edge GPU Cloud Solution to Enhance HPC Applications
Ian Buck, Vice President and Director of Nvidia's computing division, discusses the addition of a new GPU cloud solution to HPC applications: 'Previously, scientists faced numerous challenges in using these applications, such as time-consuming installations and the need to update multiple resources. Now, accessing these applications is as simple as users downloading apps from the App Store on their iPhones and iPads.
HPC and supercomputing applications are available on the NGC registration page for all users with NGC accounts. Supercomputers can run on any Nvidia Pascal system and the next-generation GPU graphics processing acceleration systems.
At the SuperComputing17 conference in Denver, Nvidia introduced these new tools to demonstrate the growing connection between the company and high-performance computing.
Over 500 HPC applications are integrated with GPU graphics processing acceleration, harnessing the power of graphics cards. According to a recent analysis report by Interesect360 Research, about 70% of the top 50 HPC applications are integrated with GPU graphics processing acceleration. From astronomy to life sciences and the medical field, GPU graphics processing acceleration applications are used in a variety of scientific disciplines. However, these applications are notoriously challenging to install.
'Applications are built from various software stacks and different libraries,' explained Buck. 'These applications are developed by researchers for their own use, not necessarily for external release.'
Nvidia has also addressed a similar issue for the deep learning and cloud community, announcing GPU Cloud for software developers last month. The new tools are now part of that subscription area.
Regarding HPC applications coming to NGC, Nvidia is starting small-scale, with five main applications: GAMESS, GROMACS, LAMMPS, NAMD, and RELION.
In terms of visual HPC, Nvidia has collaborated with ParaView to introduce three containers currently available in the beta version: ParaView with Nvidia IndeX for displaying large-scale volumetric data; ParaView with Nvidia Optix deployed for X-ray exploration, and Nvidia Holodeck providing real-time interactive imagery and high-quality visuals.
The GPU Tesla V100, based on Nvidia's Volta architecture, is now available on cloud service providers, offering cloud computing services. This includes Dell EMC, HPE, Huawei, IBM, and Lenovo, as well as cloud service providers Alibaba Cloud, Amazon Web Services, Baidu Cloud, Microsoft Azure, Oracle Cloud, and Tencent Cloud.
If you're using an NVIDIA graphics card, it's advisable to regularly update drivers for NVIDIA displays to ensure optimal support. Find details on how to update NVIDIA drivers here.
