Cutting the carbon footprint of supercomputing in scientific research

36:49
 
Kongsi
 

Manage episode 327478551 series 3008958
Oleh Physics World ditemui oleh Player FM dan komuniti kami - hak cipta dimiliki oleh penerbit, bukan Player FM, dan audio distrim secara langsung dari pelayan mereka. Tekan butang Langgan untuk mengesan kemas kini di Player FM, atau tampal suapan URL ke dalam aplikasi podcast lain.

Science benefits enormously from supercomputing, which enables researchers to process vast amounts of data and conduct complex simulations. But these machines can be notorious energy guzzlers, with the largest supercomputers consuming as much power as a small city. In this episode of the Physics World Stories podcast, scientists discuss how individuals can reduce the environmental impact of supercomputing without compromising research goals.

Simon Portegies Zwart, an astrophysicist at Leiden University in the Netherlands, says more efficient coding is vital for making computing greener. While for mathematician and physicist Loïc Lannelongue, the first step is for computer modellers to become more aware of their environmental impacts, which vary significantly depending on the energy mix of the country hosting the supercomputer. Lannelongue, who is based at the University of Cambridge, UK, has developed Green Algorithms, an online tool that enables researchers to estimate the carbon footprint of their computing projects.

Find out more on this topic in the article “The huge carbon footprint of large-scale computing“, originally published in the March issue of Physics World.

107 episod