AI > Sustainability


Byte back


AI: is innovation sustainable?


Editor Robert Roe on the critical role of computing technologies and questions surrounding the long-term sustainability of scientific computing


T


echnology and its integration with scientific computing is


leading to new breakthroughs and opening the door to new ways of conducting scientific research. This can be seen running through the themes of this issue, with AI and quantum computing taking centre stage. But how sustainable is the


current computing revolution? AI is creating many new use cases and driving large numbers of organisations, both in scientific computing and wider enterprise applications. Users want to adopt advanced computing technologies more than ever before. The sheer size of data sets and the energy consumption of hardware is exploding, leading to huge amounts of energy being used to support scientific discovery. In 2021, Dr Loïc Lannelongue,


a researcher from the University of Cambridge, published a paper entitled: “Carbon footprint, the (not so) hidden cost of high performance computing”. In this paper, Lannelongue noted that it is estimated that “data centres have a yearly carbon footprint


of 100 megatons of CO2 just


from electricity production, similar to the entire American commercial aviation sector.” While large AI systems may


be more energy-efficient than traditional HPC, the market is growing to be much larger, creating a huge demand for server technologies. More now need access to vast computing systems packed with GPUs. While at vastly different


maturity levels, AI and quantum computing technologies offer huge potential to change the way science is conducted. From automating insights and predicting the capabilities of compounds and structures to tackling problems intractable with classical systems, creating new avenues for scientific computing systems provides tools that can extend the reach of scientists and engineers. AI is delivering on the


hype promised over the past several years. The explosion in popularity of large language models (LLMs) is helping to connect predictive systems and allow further insights to be generated by scientists and engineers that can now


16 Scientific Computing World Summer 2023


ask questions to an AI system. AI has already led to several breakthrough projects such as using AI to imagine new proteins or accelerate drug discovery pipelines. To sustain this research into the future, significant changes must be made or innovations discovered that can lower the energy costs of doing business in AI. Silicon photonics offers a path to moving data more efficiently, but that is still some way off widespread adoption. Quantum can offload some applications traditionally done on HPC or AI systems, but has limited use cases for adoption. The explosion of data,


increasingly powerful computer hardware and the availability of software frameworks enable scientists who want to study and implement AI across areas such as chemistry, life sciences, astronomy, physics and materials science. Drug discovery is undergoing


a radical evolution of its capabilities due to the growing use of computational methods, including AI and ML methods. These increasingly ubiquitous approaches are driving companies to find new ways to develop candidate drug compounds and open the path towards more personalised medicine initiatives. Over the past decade, we


have begun to see AI penetrate nearly all industries and scientific disciplines, from the headline-grabbing integration


Breaking through


Later this year, we will launch a new initiative, Breakthroughs, aimed at exploring the industries and sectors that are unleashing scientific discovery with the power of computing. In print and online, it will showcase user stories and share the innovation and work being done by scientists and engineers using the latest computing technologies to accelerate their research. By revealing


the stories of how scientific computing is reshaping our world, Breakthroughs will translate technical benefits into tangible results for scientists and researchers, increasing their knowledge of available technologies and how to apply them. Watch this space!


in autonomous vehicles and the protein-folding predictions of AlphaFold, to the more quietly heralded work managing traffic flows, creating more efficient jet engines and removing noise from astronomical images. This is only the beginning, especially as AI is increasingly deployed in large systems or combined with traditional supercomputing architectures. The demand for AI is increasing,


‘Will the scrutiny around AI in the wider world become focused on our own sector?’


and so too is our reliance on the energy required to support advanced computing. The convergence of AI and


HPC promises to transform the scientific computing landscape, with the potential to enable research groups to tackle challenges otherwise beyond their capabilities. The pace of this change can be startling, and can be viewed as controversial. Will scrutiny around AI in the wider world become focused on how it is deployed in our own sector? Will we see scaremongering headlines about energy usage or sustainability? I wonder. What does the future look like for scientific computing? SCW


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42