01 Feb 2022
Explainer

# What’s next in computing: The era of accelerated discovery

To meet the growing challenges of an ever-shifting world, the ways we have discovered new ideas in the past won’t cut it moving forward. A convergence of computing revolutions taking place right now will help accelerate the rate of scientific discovery like nothing before.

To meet the growing challenges of an ever-shifting world, the ways we have discovered new ideas in the past won’t cut it moving forward. A convergence of computing revolutions taking place right now will help accelerate the rate of scientific discovery like nothing before.

The last two years have been like few that humanity has endured in generations. But they’ve also shown us the value that scientific discovery can bring to the world: Scientists from across the globe came together to synthesize and develop vaccines for COVID-19 in a matter of months. That’s a step-change from the 10 to 15 years it typically takes for vaccine and drug discovery, thanks in part to capabilities of modern computing.

Computing has never mattered more to the world than it does now, and it’s allowing researchers to super-charge the scientific method — humanity’s most powerful and successful process for advancing our knowledge and solving complex problems.

We’re at a moment in time where computing is poised to reshape just about every industry, field of study — and even society itself. Across the fields of AI, cloud, and quantum computing, we’ve seen massive advancements in recent years. It’s the result of decades of research, scientific curiosity, and plain hard work, across a variety of scientific and technical disciplines of which we’re now starting to see the culmination.

AI has emerged as a driving force behind countless automated business decisions every day. Cloud computing has expanded into a set of systems that has put us on a path towards being able to leverage the entire world’s compute power as if it were a single infinitely powerful computer. And quantum computing has evolved from a theoretical concept discussed at academic conferences to a field of study that has the potential to upend industries the world over.

Separately, these three technologies are producing new solutions to problems that are changing the world every day. But when we’re able to leverage some of them, or all three together, in new ways, the potential for discovering new materials, finding novel drugs and new uses for existing ones, solving science’s hardest problems, and myriad other revelations, will be unlike anything we’ve seen before.

We’ve come to call this accelerated discovery. It’s now a core part of the work we’re doing at IBM Research. We believe, given all these advancements in computing in recent years, and our deep research across these fields, we’re in a unique position to leverage accelerated discovery for IBM, our partners, and the world at large. The world is changing rapidly every day. Science and computing are critical to tackling the urgent problems society is wrestling with, from fighting a pandemic to reversing the impact of climate change.

Our efforts in accelerated discovery aim to develop, validate, and incubate technologies that accelerate the implementation of the scientific method, harnessing the capabilities of AI, hybrid cloud, and quantum computing. We need these cutting-edge technologies, working together, to address society’s greatest challenges — to find novel sustainable materials able to help us fight challenges from pandemics to climate change and so much more.

With the innovations on the horizon through these computing revolutions, we suspect that we’ll be able to accelerate discoveries up to 10 times faster and with 10 times less cost than they’ve been done in the past. This will allow us to help uncover breakthroughs that will affect everything from our daily lives to corporate innovation and government policymaking.

We’ll be able to accelerate discoveries up to 10 times faster than they’ve been done in the past.

At IBM, we’ve been researching all three technologies for decades, and are working on foundational infrastructure that underpins all of this work, from our quantum processors, to more efficient AI processors, and new processor architectures that could keep Moore’s Law alive for years to come.

We’ve started considering an even brighter future, where all three of these revolutions can work together to create something even more astounding. What can be achieved with an AI algorithm that has the reach of the entire cloud behind it, as well as the processing ability that quantum computing could unlock? What can be uncovered, and at what speed, when these three revolutions are working together? This is what we mean by “accelerated discovery,” and we believe that the combined progress of these three fields of computing will shape the world like few things have before.

## Accelerating discoveries with our partners in the wider scientific community

Last year, we launched our first Discovery Accelerators, which are hubs for accelerating scientific breakthroughs in specific fields of research, with help from academic, governmental, and industry partners.

The Discovery Accelerator we established in collaboration with Cleveland Clinic, one of the preeminent health care institutions in the United States, is focused on advancing the pace of discovery of materials and treatments useful in health care and life sciences through the use of AI, hybrid cloud and quantum computing technologies. As part of this collaboration, we’ll install our first private-sector IBM Quantum System One in the U.S. on Cleveland Clinic’s campus.

In health care, it takes an average of 17 years between knowing on a scientific level that something can work, to the time when it’s available for use. There are hundreds of steps researchers need to carry out for just a single project, and the months they can take soon add up to years. With our partnership, we’re aiming to assist Cleveland Clinic in dramatically accelerating the discovery workflow.

Last summer, we launched the Hartree National Centre for Digital Innovation (HNCDI) with the UK’s Science and Technology Facilities Council. The goal for this Discovery Accelerator is to solve some of the most ambitious problems in science, business, and society, deploying the next frontier of high-performance computing, AI, quantum computing, and hybrid cloud from IBM Research. The center is fueling collaborations across industries and research ecosystems in fields as diverse as life sciences, environmental sustainability, and manufacturing. Look out for early results of research conducted at the HNCDI soon.

These are just the first seeds that we have sown, and while there are many more to follow, we’re already seeing the value in the technologies we’ve created. We built RXN, an AI tool that can predict the most likely pathway for a chemical reaction and understand natural-language descriptions of chemical procedures. Thieme, one of the world’s leading medical publishers, recently began using RXN with its own datasets to accelerate discoveries in organic chemistry.

We’ve also seen how ideas generated by AI can help us come to discoveries quicker. Researchers from IBM, Oxford, Cambridge, and the National Physical Laboratory, showed last summer how AI-designed antimicrobial peptides interact with computational models of a cellular membrane that could potentially have wide implications for drug discovery and delivery.1 Our researchers are also using quantum computers, AI, and hybrid cloud to explore climate change-mitigation solutions, such as materials to efficiently capture carbon and resources to compute more sustainably.

We’ve even started to see just how valuable using technology to accelerate discoveries can be through our own research projects. In the field of materials discovery, identifying a new molecule usually takes 10 years of research and can cost between $10 million and$100 million to develop. With Project Photoresist, which we showed off last year, we were able to shrink that process down to months. Using a combination of AI and robotics systems, we were able to uncover, simulate, and test a new photoacid generator, which is used in lithography, a key process in developing computer chips. We were able to synthesize the new material 100 times faster than more traditional materials discovery methods, and we hope to use our processes to build out similar discoveries of our own, and our partners, this year.

If the last few years have shown us anything, this step change in the convergence of computing technologies couldn’t have come at a better time. The world is changing rapidly every day. We need the speed that accelerated discovery can bring us to uncover solutions to the world’s most complex problems. The traditional timeframes for research and development are no longer satisfactory for our increasingly interconnected world.

Without accelerating the rate at which we discover solutions, there is no path to be able to address our most pressing challenges. This is not just about working faster, it's a completely different paradigm. An accelerated discovery mindset can benefit organizations of any kind.

Science is a creative process. It's about imagination, it's about being able to navigate a discovery space, or a design space that is practically infinitely vast. It requires a flash of insight, some imagination, some creativity, the ability to land on something novel, maybe surprising, even. That discovery mindset can be used in new areas of problems that might have seemingly little to do with science on the surface.

01 Feb 2022

## References

1. Das, P., Sercu, T., Wadhawan, K. et al. Accelerated antimicrobial discovery via deep generative models and molecular dynamics simulations. Nature Biomedical Engineering (2021).