Once fully operational in 2021, Aurora Argonne's Aurora Laboratory will bring unprecedented levels of performance to t exaflop range, billion billion (1018) Calculations per second. Scientists are similar to Dr.. William Tang, Principal Research Physicist at Princeton Plasma Physics Laboratory, ready to take full advantage of the scientific potential of the previous scientific efforts of the system.
Tang and his team seek new approaches fusion reactions to the generation of electricity, enabling abundant energy for the growing world population. Fusion is the kind of power that the sun and stars produce.
“Clean energy, delivered on a huge scale, would free our imagination to explore new ideas and approaches. However, if we want to deliver clean energy to the world, they need computers to accelerate scientific progress and insights, ”said Tang.
As part of the Argonne Leadership Computing Facility Early Science Program, Tang is leading a project that will use deep learning methods and artificial intelligence to promote predictive capabilities for fusion energy research in the exascale era.
“As a result of Stephen Hawking's remarkable mindset, there was an important fusion for the future energy needs of humanity. Of course, fusion occurs in nature. However, creating it in an earthly environment is a major challenge, ”said Tang. “Climate change is a major challenge for our planet. Not only is there a reduction or elimination of urgent carbon emissions; it is vital. The energy of the future comes from clean and safe fusion. We face significant challenges in making this transition. But today, it is an achievable goal, thanks to exascale computing, the emergence of AI, and deep learning. ”
Tang's vision for fusion-based energy brings many benefits over today's nuclear power stations. Since a fuel-containing isotropic hydrogen isotope of less than one-minute hydrogen is derived from marine water and tritium bred in the facility – in the reaction room, fusion systems cannot cause disaster or explosion. Also, because the radioactivity that creates the fusion process resides, there is no risk of long-term environmental pollution arising from the solution.
Keeping the gene in the bottle
It is extremely difficult to replicate the science of Mother Nature. The fusion process within our star – the Sun – comes to notice in plasma temperatures in the tens of millions of degrees. Future fusion facilities must create warming which is much warmer to produce fusion reactions. For this reason, the approach to using physical barriers to maintain the plasma is impractical. Most materials are destroyed when exposed to such temperature extremes, so the retention effort requires innovative methods.
“We have invested a lot of money to deliver clean fusion energy through childbirth methods,” indicated Tang. “However, there are many obstacles to be overcome. One big challenge is to make a quick and accurate prediction of turbulent adventures' which allows hot thermal plasma to escape quickly. Supervised mental learning helps us as a predictive guide. If we can predict that we will come to an "accident", we can plan to control it. "
Such physical science consists of extensive data sets. Optimized optimization networks that support the Tang project need to interpret data that reflects the three-dimensional space, plus the fourth element, time. The challenge, therefore, is the ideal approach to training the system to follow a logical pattern when handling large amounts of data.
“Computer computers are a major development in the way we carry out calculations. In ancient times, abacus did the job. In recent years, sliding rules, calculators, and more powerful computers have progressed in science in significant ways, ”said Tang. “However, with computing at exascale level, we have new ways of tackling major challenges that require very fast and accurate calculations. With extremely powerful systems like this on top, we can open our imagination with new possibilities that are considered practical or impossible but five years ago. Compared to the traditional approaches we use as benchmarks, it is exciting how fast we can make progress today. ”
Computing on Computing
It is an effort to build a system on the Aurora scale, which requires government funding assistance to bring the latest hardware and software together into a huge one-way system. To achieve the level of performance required by modern science, a new generation of Intel Xeon scalable processors and Intel Optane DC Memory will, together with Intel X's future as part of Aurora's system, built by Cray.e technologies.
“These industry-laboratory collaborations are critical to developing a system that will enable innovative science, and encourage the best and brightest young people around the world to engage in critical research efforts,” said Tang. “With today's scientists and today's scientists developing new technologies and AI, we can make innovative progress and accelerate the pace of our ultimate goal, providing something that is vital for the nation. human. ”
Validation of Science Theories
Often in research, it is extremely difficult to observe theories in a global environment. The pursuit of the Higgs boson, or the attempt to identify the gravitational waves predicted by Albert Einstein's general relationship theory, gives two such examples. In each of these cases, scientists adopted the reality of the phenomenon of her theoretically level for several years. However, the validation of these theories involves detailed experimental observations.
Added Tang revealed, “Exascale's computing ability to deal with much larger amounts of data enables us to create what was never possible. The incredible speed of major computers also reduces the time to discovery. It now takes hours or days now work has been months or years now. Therefore, finally, we have the ability to validate statistically theories and create their reality. We are very excited to be part of the selected team for the first exascale computer exercise. ”
The road ahead
While the Tang team focuses on a new approach to clean energy, Aurora will also support progress in other scientific disciplines such as climate monitoring, cancer research, and chemistry. “Aurora's capable architecture is a new architecture, but it has created proven technologies. Going on new research roads with new tools, the right training is always important, ”he said. “However, we feel confident that we have the experience to face new challenges ahead. We must be adaptable as scientists. At the moment, we're excited about going on to the next stage. ”
Tang speaks optimistically about the very complex work he and other researchers will do with Aurora. “My work is made possible by technological advances in the 21st Century. AI has been around for some time, but accelerated development of neurological nets and other methodologies that enable exascale computers allow us to use more efficiently.
The team cannot fully replicate what Mother Nature already does, but as Tang puts it, “Greater communication power gets closer to us. Tomorrow's exascale systems and the new insights gained from them will give us the power to do more amazing things in the years ahead. Our exciting work is both exciting and exciting because we have the opportunity to do something that will benefit the world. ”
Rob Johnson spent much of his professional career consultation for Fortune 25 technology company. Rob currently owns Rob Fine Tuning, LLC, a strategic consulting and marketing company based in Portland, Oregon. As a technology technician, sound, and excellence in his enthusiastic life, Rob also writes for TONEAudio Magazine, conducting reviews of high-end home audio equipment.
This article was produced as part of the HPC Intel editorial program, with the aim of highlighting cutting edge science, research and innovation driven by the HPC community through advanced technology. The publisher of the subject has final editorial rights and determines the articles published.