How quickly is climate change happening? What is the effect of human activity on the environment? Was the recent devastation caused by Hurricane Harvey made worse by global warming? These are just a few of the profoundly complex questions that scientists are attempting to answer using massive amounts of climate data and powerful high performance computing (HPC) tools.
It’s difficult to deny that climate change is one of the grand challenges currently facing humanity. Former President Obama even labeled it the greatest threat to future generations, larger than even terrorism, economic instability, or disease. It is these grand challenges that drive researchers to consistently bolster computing capabilities, and make huge investments in new tools and technologies that promise to help resolve these tremendous issues.
Addressing the grand challenge of climate change is dependent on the ability to quickly uncover patterns in large amounts of data to arrive at insights that will help protect lives and property, offer advance warnings of extreme weather events, and prepare for the implications of a rapidly changing environment. The HPC industry has realized that these activities require a new generation of supercomputers, and as they work to build these systems, they are considering the human brain as an ideal model.
The brain is arguably the world’s most well-made computer – it’s small enough to fit into a shoebox, is more efficient and economical to operate than any hardware in existence, and can process and make sense of many different types of inputs simultaneously. Researchers are investigating ways to develop computing systems that solve problems in the same way the human brain does, by learning and adapting based on the introduction of new data, and thinking, reasoning, and remembering using vast quantities of information.
The result is the emerging field of cognitive computing, a sub-field of artificial intelligence (AI) which simulates the human thought process via computerized models. This type of computing is still evolving but is showing great promise as a way to rapidly solve complex, multi-layered, and data-heavy problems. In the weather and climate industry, researchers are incorporating these techniques to train a new generation of computers that may be able to autonomously uncover new weather patterns, predict extreme weather events farther into the future, and help us quantify how climate change might evolve.
According to a new report from Grand View Research, the global cognitive computing market is expected to grow to just over $49 billion by 2025, driven by companies in a variety of industries needing to analyze large volumes of data optimally on a real-time basis.
Source: Grand View Research, December 2016
Cognitive computing is made possible with parallel processing capabilities, which enables large computational problems to be divided into smaller tasks and executed simultaneously – much like the way the human brain breaks up and solves problems. Graphics processing units (GPUs) like those from NVIDIA are becoming the processor of choice for these specialized HPC tasks, as the parallel processing capabilities of NVIDIA GPUs make them more capable of handling large volumes of complex datasets and quickly solving data-intensive problems. To bring these capabilities to a broader range of HPC customers, Hewlett Packard Enterprise (HPE) offers a selection of GPU-optimized hardware solutions that can help companies rapidly adopt and integrate NVIDIA GPU technologies.
HPE has helped lead the industry in harnessing cognitive capabilities. With the HPE Deep Learning Cookbook, customers have access to a massive collection of performance results utilizing various hardware/software stacks and analytical performance models. This valuable resource enables users to effectively plan, deploy, and leverage cognitive computing solution to achieve deep insights.
-wide investigation into alternative forms of computing for years. Last April, HPE introduced a public, open-source version of the platform called the Cognitive Computing Toolkit, which leverages NVIDIA GPUs to provide a powerful software framework for building massively-parallel applications, like many of those commonly used in weather and climate modeling.
As we work to address the grand challenge of climate change, researchers and data scientists are increasingly adapting their understanding of how the brain works to fuel the next generation of computing systems. HPE and NVIDIA are providing a portfolio of GPU-accelerated compute solutions to power these critical activities, and simplifying the adoption of GPU computing among companies in nearly every industry and vertical application.
For more news and information about how your business can exploit cognitive computing applications to improve business processes, please follow me on Twitter at @pango. You can also check out @HPE_HPC and @NvidiaAI for additional information.