COVID-9 High Performance Computing (HPC) Consortium

  • During the pandemic, the COVID-19 High Performance Computing (HPC) Consortium made the world’s most powerful computers available for free to researchers trying to fight COVID-19.
  • COVID-19 showed the power of global collaboration and the need to tackle global threats ahead.
  • The creation of a special global consortium led by peer-nominated experts, would assess and respond to future global risks and emergencies such as droughts, earthquakes, asteroids or other deadly diseases.

In March, COVID-19 had hit Asia, Europe and the US, and the world was gripped by disbelief. The virus was far deadlier and spread much faster than any disease we’d ever had to address in modern times. Fighting back against it, we were surprised once more as we developed new capabilities and new partnerships faster than ever before. Some of these initiatives could shape how global collaborations address risk in our post-COVID future.

We must apply what we’ve learned during the pandemic on a wider scale and establish a new international organization: the Science Readiness Reserves. It would assess risks and help prepare for future global emergencies – from another deadly virus to a super vulcano eruption or a nuclear power plant disaster, and much more. This way once a disaster strikes, we won’t be scratching our heads in panic; we will be ready to act and save lives.

What is the World Economic Forum doing about the coronavirus outbreak?

How a global consortium – established quickly – is fighting COVID-19

One successful current initiative is the COVID-19 High Performance Computing (HPC) Consortium. A U.S. organization with an international dimension, it’s offering some of the world’s most powerful computers, for free, to researchers trying to fight COVID-19. More than 60 teams of scientists are already relying on them, with new researchers joining weekly. The Consortium’s members collaborate with the Partnership for Advanced Computing in Europe (PRACE), an organization that unites supercomputing partners on the other side of the Atlantic. Researchers from both bodies regularly exchange data, ideas and results.

The history of Supercomputing at IBM
The history of Supercomputing at IBMImage: IBM

Being a public-private partnership with government support, the Consortium solves key problems for researchers racing against time. To get results quickly, modern science often needs a computing boost – to sift through thousands of potential designs of molecules when developing a new drug, or to simulate the spread of the virus in its second or third waves. Typically, to get computing power on a supercomputer, researchers submit a proposal and wait weeks or even months just for access. And it needs support of policymakers to zoom past any typical bureaucratic hurdles.

The COVID-19 HPC Consortium is providing exactly that. One team, led by Amanda Randles, a biomedical engineer at Duke University, has developed airflow simulations for a new device to split a ventilator between two or more patients. For its work, the team needed more than 500,000 compute hours; through the Consortium, they got matched with a specific supercomputer, and completed research that might have taken months in just six weeks. The scientists have just submitted their results to the FDA for approval.

Randles’ team’s result is not the only success story of the Consortium. Researchers at British AI startup PostEra used a different supercomputer to build a crowdsourcing platform for scientists worldwide to submit drug designs. The team received more than 5,000 submissions and managed to process them within just a few days – something they wouldn’t have been able to do without the Consortium. Over the past two months, they’ve narrowed the number of designs to around 200 and tested them in the lab. They’ve discovered that 21 molecules seem to be effective against the targets in the key protein within the coronavirus. Their next step will be to make sure the drug is fit for purpose before it can go into humans, the most important part of the drug discovery cycle.

Have you read?

How collaboration could aid in future crises

Impressively, it took less than a week to launch the Consortium and for all partners to agree on the ground rules. We quickly set up the proposal review and matching committees to approve and find the best supercomputing partner for each team. The Consortium is the proof that we know how to quickly mobilize the world’s talents to solve global problems. I’m certain that it could be done again, on a much wider scale, and applied to future emergencies.

Together with the Chair of Harvard Astronomy department Avi Loeb, I’ve recently voiced my thoughts on just that – how to expand the idea of the Consortium to create a new body, the Science Readiness Reserves. This group, led by peer-nominated experts, would assess and respond to future global risks and emergencies. It would be run by scientists, experts in a variety of fields able to address the consequences of an asteroid impact, a mega drought, a powerful earthquake, a volcano eruption or another, even deadlier, disease.

“We know how to quickly mobilize the world’s talents to solve global problems. I’m certain that it could be done again, on a much wider scale, and applied to future emergencies.”—Dario Gil

Such an organization could unlock technological capability. Critically, this body would ensure that scientists always have access to technologies they need (such as a supercomputer or perhaps soon even a quantum computer) otherwise difficult to obtain to deal with and prevent future emergencies. The aim would be to assess risks and put together a set of specific scientific capabilities that would be of value in the future, along with recommendations for the government, the health systems, emergency services, economists, other researchers, the media and the public.

Key to SRR’s success would be matchups between the research team submitting a proposal and other scientists in academia or industry, or organizations in a specific field (say, the World Health Organization in the case of a pandemic). Such strategic partnerships between industry, academia and government could dramatically speed up research and development and be key for promoting creativity through the exchange of data and ideas.

Today, scientific talent, expertise, and resources are highly distributed across government agencies, academia and the private sector. Still, global and national bodies deal with specific fields – again, take the WHO, or UN, or NASA – and their core priority is not solely to assess future risks. And while many groups and movements advocate for wide dissemination of research and greater global collaboration, the SRR would specialize in ensuring that key technologies and expertise are made available to prevent global crises.

Such an organization does not currently exist – despite the fact there is no shortage of global emergencies, from climate change to antibiotic resistance to epidemics and pandemics. It might be that before the Consortium, we didn’t think industry rivals would be that keen to collaborate. Or it might be because we thought that having separate bodies each targeting a specific global threat is more efficient. Or because it is only recently that technology has matured to the point of anyone can access a supercomputer on the other side of the world through the cloud as if it were standing in their own lab. I think it’s all of the above – but what’s clear is that now we have all the ingredients to make it happen.

We need to act – now. Policymakers must grasp the importance of such an organization, create the necessary regulatory frameworks and establish direct links with the future SRR experts. We must embrace the open-source approach to software and data sharing.

As we do, we must create the necessary defenses against cyberattacks and other criminals. COVID-19 brought shocking moments in research, including hackers extorting money from the University of California, San Francisco in exchange for COVID-19 research data. Such vulnerabilities are unacceptable. We have to make sure we have necessary protocols in place to protect the work of SRR experts on future ‘known unknowns’ so that we can act on their recommendations and design new drugs or materials promptly and efficiently.

I know we can do it. The HPC Consortium is the case in point to prove it. The COVID-19 crisis has shown us all how unprepared we were to fight this global challenge. Thankfully, the crisis has also taught us we have the capability to make great strides once we pool our resources and talents. Realizations like that will boost our chances to survive this and other crises to come.

(ref)