“Thanks to massive mobilisation involving thousands of individuals and organisations, Folding@home has exceeded the symbolic threshold of the exaflop.”
In March 2020, distributed computing project Folding@home exceeded the symbolic threshold of the exaflop and reached a computing speed superior to that of IBM’s Summit, the most powerful supercomputer at the time (since superseded by Fujitsu’s Fugaku). This feat was made possible thanks to the mobilisation of thousands of internet users, businesses and public institutions, who decided to share their computers’ resources in the context of the COVID-19 pandemic so as to enable modelling of the proteins responsible for the disease and accelerate discovery of a treatment.
Scientific research requires phenomenal computing speed to carry out ever more complex simulations, which are essential in the solving of large-scale problems in different areas. However, researchers do not always have the financial means that would enable them to use a supercomputer.
By “distributing” a computing calculation across thousands of personal computers connected to the internet, shared computing works like a super powerful virtual computer that can solve large-scale computing problems, which would otherwise require a very long time in a traditional environment. The users download and install a piece of software onto their machines. This software runs in the background, only using available and unused resources to perform the calculation tasks, it then sends the results back to the source. Most of the time, each user can choose the project or projects in which they wish to take part.
From the search for signs of extra-terrestrial life…
The pioneer in the field is the SETI@home project, developed by the University of California, Berkeley, in cooperation with the SETI (Search for Extraterrestrial Intelligence) programme. Launched in 1999, it had two objectives: to prove the viability of distributed computing on the one hand, and to detect the presence of extra-terrestrial life by analysing data provided by the Arecibo radio telescope (on the island of Puerto Rico) to spot any signals on the other.
Only the first objective was achieved. SETI@home proved to the scientific community that distributed computing made it possible to obtain computing power that could compete with the most powerful supercomputers.
The BOINC platform, which has supported the software since 2005, now supports several distributed computing projects in a wide range of areas: biology and medicine, physics and nanotechnologies, astronomy, climatology, mathematics and computing. It brings together a large community of users and reaches a total average computing power of over 33 petaflops spread across around 550,000 computers.
In 2019, the two most popular projects on the platform were Collatz Conjecture, which aims to refute the eponymous statement by testing mathematical sequences, and Einstein@Home, destined to detect gravitational waves by analysing interferometer data.
… to the fight against diseases
In parallel, BOINC is used in the World Community Grid (WCG) project. Created in partnership with IBM in 2004, WCG brings together several scientific research projects of strong humanitarian interest – the fight against AIDS or the Ebola virus, climate study, research for new materials in the renewable energy field, etc. – in a unique computing grid (users are included in all projects by default).
Among these projects, Décrypthon has facilitated progress in the understanding of genetic and rare diseases. Launched during the 2001 Téléthon by the French Muscular Dystrophy Association (AFM) and IBM, the project aimed to produce a proteome map (the entire set of proteins expressed in a cell), to be made available to researchers. In total, 75,000 internet users got involved.
Each computer contributed approximately 133 hours, that is over 10 million hours computing in total, and 550,000 proteins of the living world have been catalogued. It would have taken over 1,170 years to do this using only one computer. Following this, around ten scientific projects selected through calls for proposals have been carried out within the framework of the Décrypthon programme.
Also in the area of research on proteins, Folding@home, founded in 2000 by Vijay Pande at Stanford University in California, seeks to further the understanding of protein folding by simulating this process in diverse conditions. The aim being to draw useful knowledge that can enable the development of new medications against several diseases, such as Alzheimer’s disease and certain types of cancer. The computation results will be available to scientists the world over.
In February 2020, the Folding@home team announced that part of its efforts would be focused on research into a treatment against SARS-CoV-2, the virus at the origin of COVID-19, and launched an appeal to internet users encouraging them to contribute from their home by installing the software.
The aim is to understand how the viral proteins responsible for the disease work (read below “How the proteins responsible for SARS-CoV-2 work”) by 3D-modelling them.
Massive mobilisation involving thousands of individuals and organisations has enabled Folding@home to exceed the symbolic threshold of the exaflop. Over the following months, the researchers put this computing power to use: “We’ve simulated nearly the entire proteome of the virus and discovered more than fifty new and novel targets to aid in the design of antivirals”, declared a researcher from the Washington University School of Medicine . “We have also been simulating drug candidates in known targets, screening over 50,000 compounds to identify 300 drug candidates.”