Huerta’s innovation will help observe the cosmos more clearly at space
Black holes are always a hot topic among astronauts. One of the reasons behind it is that black holes are unpredictable and far away, making researchers wonder what they can do. However, in recent years, Artificial Intelligence (AI) features have helped scientists figure out answers using technology for their confusion.
Since the discovery of phosphine gas in the clouds of Venus, the planet that is incapable of housing living beings because of its high temperature, the intensity of expectation on space has widened drastically. Scientists are looking at new possibilities in the expanded no ending topic. The ancient stories of aliens coming to earth in weird space vehicles are still attractive to everyone. Henceforth, AI is aiding researchers to zoom in more at the vast black space. Their focus is heavily on black holes that never seem to conclude.
Black holes are points in space that are so dense to create deep gravity sinks. They even resist light to pass through its surface. Anything that goes too near the black hole, whether it is a star, planet, or spacecraft, will be stretched and compressed like putty in a theoretical process aptly known as spaghettification. If humans go into black holes, they are predicted to be shattered into pieces.
Albert Einstein first predicted the existence of black holes in 1916 with his theory of relativity. Almost after five decades, black holes got its name. Since then, scientists and researchers are looking at ways to deepen their research on black holes. The new trending way to learn about black holes is by suing the technological features of artificial intelligence. It is not just black holes unraveled by AI; other space-related topics are also interconnecting with AI for better discovery.
Huerto’s initiative to merge AI and high-performance computing (HPC)
Eliu Huerta is a scientist from the National Centre for Supercomputing Applications and Department of Astronomy, University of Illinois. He has merged AI and high-performance computing (HPC) to observe the cosmos more clearly at space. For several years, astrophysics researchers used data to detect signals produced by collisions of black holes ad neutron stars. Huerta’s research, if succeeded, will be a breakthrough to the space discoveries. It is expected to answer a thousand unanswered and prolonged questions.
However, the debates on space-time fabric ended in 2015 when researchers observed the gravitational waves for the first time. They used pattern-matching techniques on data from the Laser Interferometer Gravitational-Wave Observatory (LIGO), home to some of the most sensitive instruments in science.
Discovering black holes using disruptive technology
Recently, astronomers announced that they had detected a massive collision of two black holes ever discovered. The clash is expected to have happened seven billion years ago, leaving only the late signs to humans. The cataclysmic event offered researchers a front-row seat to the birth of one of the Universe’s most elusive objects.
A team of researchers in 2018 used an AI bot named ‘ClaRAN’ that scans images taken by radio telescopes, which could spot the galaxies that emit powerful radio jets from supermassive black holes at their centers. It was then discovered by data specialist Dr. Cheu Wu and astrologer Dr. Ivy Wong from the University of Western Australia node of the International Centre for Radio Astronomy Research (ICRAR).
Ultimately, the new technology to discover space-related information in the trending 2020 is supercomputers that process the data that instruments gather in a single day. In 2017, Huerta’s team showed how a deep neural network running on an NVIDIA GPU could find gravitational waves with the same accuracy in a fraction of the time. The AI model Huerta used was based on data from tens of thousands of waveforms. He trained it on a single NVIDIA GPU in less than three hours.
Having a binocular view at black hole spin
Huerta and two of his students created a more sophisticated neural network that can detect the motion of two colliding black hole’s spin. The AI model accurately measured the faint signals of a small black hole merging with a larger one.
The trio used an IBM POWER9-based system with 64 NVIDIA V100 Tensor Core GPUs that took 12 hours to train the resulting neural network with a requirement of 1.5 million waveforms. To further the process, the Huerto team got access to 1,536 V100 GPUs on 256 nodes of the IBM AC922 Summit supercomputer at Oak Ridge National Laboratory. By making a connection between Summit’s GPUs and its IBM POWER9CPUs, they trained the model in 1.2 hours. This is quite an achievement in the evolving outer space stories. The result was published in ‘A paper in Physics Letters B.’ It showed the combination of AI and HPC, which solved challenges in astrophysics.
The innovation is marking a milestone in Artificial Intelligence’s extensive features and its functions on space-related topics. Both space and AI are something that induces curiosity. It will be interesting to see the emerging discoveries from space and AI gadgets or technologies that feature the discoveries.