Fighting COVID-19 in New Era of Scientific Computing


Scientists and researchers around the sphere are racing to search out a remedy for COVID-19.

That’s made the work of all those digitally gathered for this week’s excessive performance computing convention, ISC 2020 Digital, extra vital than ever.

And the work of those researchers is broadening to embody a wider fluctuate of approaches than ever.

The NVIDIA scientific computing platform plays an critical characteristic, accelerating growth across this entire spectrum of approaches — from details analytics to simulation and visualization to AI to edge processing.

Some highlights:

  • In genomics, Oxford Nanopore Applied sciences used to be in a pickle to sequence the virus genome in precisely 7 hours the usage of our GPUs.
  • In infection evaluation and prediction, the NVIDIA RAPIDS crew has GPU-accelerated Plotly’s Flee, an details visualization instrument, enabling clearer insights into actual-time infection payment evaluation.
  • In structural biology, the U.S. National Institutes of Neatly being and the University of Texas, Austin, are the usage of GPU-accelerated tool CryoSPARC to reconstruct the predominant 3D constructing of the virus protein the usage of cryogenic electron microscopy.
  • In therapy, NVIDIA labored with the National Institutes of Neatly being and built an AI to precisely classify COVID-19 infection basically basically basically based on lung scans so atmosphere suitable therapy plans may possibly moreover moreover be devised.
  • In drug discovery, Oak Ridge National Laboratory ran the Scripps Research Institute’s AutoDock on the GPU accelerated Summit Supercomputer to mask a billion attainable drug combinations in precisely 12 hours.

  • In robotics, startup Kiwi is constructing robots to narrate clinical supplies autonomously.
  • And in edge detection, Whiteboard Coordinator Inc. built an AI machine to automatically measure and mask elevated body temperatures, screening effectively over 2,000 healthcare workers per hour.

It’s no doubt inspirational to collect up every single day and peek the incredible effort going on around the sphere and the characteristic NVIDIA’s scientific computing platform plays in helping realize the virus and discovering attempting out and therapy ideas to fight the COVID-19 pandemic.

The cause we’re in a pickle to play a characteristic in so many efforts, across so many areas, is thanks to our solid take care of offering end-to-end workflows for the scientific computing community.

NVIDIA scientific computing platform

We’re in a pickle to produce these workflows thanks to our formula to plump-stack innovation to high-tail all key application areas.

For details analytics, we high-tail the predominant frameworks admire Spark3.0, RAPIDS and Dask. This acceleration is built the usage of our domain-particular CUDA-X libraries for details analytics equivalent to cuDF, cuML and cuGRAPH, along with I/O acceleration applied sciences from Magnum IO.

These libraries hold millions of lines of code and present seamless acceleration to builders and users, whether or not they’re setting up features on the desktops accelerated with our GPUs or operating them in details facilities, in edge computer programs, in supercomputers, or within the cloud.

Similarly, we high-tail over 700 HPC features, in conjunction with the full most in overall former scientific features.

NVIDIA quickens all frameworks for AI, which has turn into the largest for duties the place the guidelines is incomplete — the place there must not any first suggestions to work with or the predominant precept-basically basically basically based approaches are too slack.

And, attributable to our roots in visible computing, NVIDIA supplies accelerated visualization solutions, so terabytes of details may possibly moreover moreover be visualized.

NASA, as an instance, former our acceleration stack to visualize the touchdown of the predominant manned mission to Mars, in what’s the sphere’s largest actual-time, interactive volumetric visualization (150TB).

Our deep domain libraries also present a seamless performance boost to scientific computing users on their features across the diversified generations of our structure. Going from Volta to Ampere, as an instance.

NVIDIA’s also making all our original and improved GPU-optimized scientific computing features readily available via NGC for researchers to high-tail their time to insight

Collectively, all of those pillars of scientific computing — simulation, AI and details analytics , edge streaming and visualization workflows — are key to tackling the challenges of as we bid time, and the next day to come.

Leave a comment

Your email address will not be published.