Investing in Advanced Research Computing resources is essential to competitive science and innovation in Canada
By Mark Dietrich, President and Chief Executive Officer, Compute Canada
In the world of science, we already know that many of the problems we face are already too big for any one institution, industry, province, or even nation to solve alone. The scientific community has addressed this problem for decades, and we have seen investments in a growing number of “big science” projects. These are massive national and international collaborations, many of which have already improved lives, advanced economies and achieved stunning scientific results.
Advanced Research Computing (ARC) powers dynamic invention and innovation in almost every sector of Canadian industry — from curing disease, to aerospace and transportation, to manufacturing and consumer goods. ARC has transformed how the world conducts scientific and engineering research, enabling discovery, insight and development in ways we once thought were impossible. These machines and the experts that support them are key to extracting value from big data, and underpin development of a diverse and well-prepared 21st century workforce. Many times more powerful than your desktop computer, supercomputers are essential national infrastructure for innovation and research.
Compute Canada’s national integrated platform of scientific computing resources, research data management tools, data cyberinfrastructure, software and expert support are available to all Canadian researchers—more than 10,000 of whom already use our resources today.
Solving Grand Challenges frequently requires tremendous computer power. Computer simulation can short-circuit years of laboratory experimentation. The more computing power you throw at problems, the more accurate the results and the faster the scientific progress. Instead of predicting weather 30 minutes from now, we can predict weather 30 days from now. Instead of finding a cure for cancer in our grandchildren’s lifetime, we can possibly get there in our lifetime.
ARC is clearly an essential ingredient in science excellence. A 2016 Compute Canada report study found that the average Field Weighted Citation Index (FWCI) for publications that made use of its computing powers was substantially above the world average in the 22 disciplines where more than 100 Compute Canada-enabled papers were reported. In some cases, it was more than double. While the Canadian FWCI average tends to be higher than the world average, Compute Canada-enabled publications also consistently exceeded this higher domestic score.
There are challenges. New scientific methods and instruments, such as genomic sequencers, are generating orders of magnitude more data than was available even five years ago. No one research institution or consortium can muster the ARC resources necessary to analyze the data at hand. Pooling and sharing ARC resources is the only way for Canadian researchers to make significant headway on big problems.
Compute Canada’s new systems, coming online this year and next, will take our collective computing capabilities from 2 petaflops to 12 petaflops and vastly enhance our data storage and management capabilities. This welcome investment is replacing aging capacity but it still will not allow us to keep pace with the science requirements in Canada.
The risk to Canada from falling behind is very real. Compute Canada is currently able to meet only half the advanced research computing requirements in Canada. We know that, based on estimates of the researchers we serve today, computational requirements will increase 7x over the next five years, and storage requirements will be see a 15x increase.
Advanced research computing is one of those scientific projects that is too expensive for any one institution or province to shoulder. In the United States, XSEDE is a nationwide project that integrates the operations of large high performance computing centres, and dozens of smaller centres, into a seamless facility that can be accessed from across the country. In the European Union, PRACE (Partnership for Advanced Computing in Europe) and EGI (the European Grid Infrastructure) were created to achieve the same results across Europe for users of supercomputing and high-throughput computing systems. All three organizations are federations, federations that have spent years building consensus around operating procedures, common systems and performance expectations, creating more value for their users than if each partner were operating independently.
As Canadians, we care deeply about maintaining global stature in core scientific areas, and over the last few years we have been falling behind. Today, we are at risk of losing researchers and ARC experts to other countries with the capacity to provide them with the state of the art advanced computing resources and are at risk of losing more. Compute Canada has analyzed public data on high performance computing systems around the world (the “Top 500” list, top500.org), aggregating computational capabilities found in the academic and research sectors by country, and comparing the total computational capability with the total number of researchers in higher education, as reported by UNESCO. Using the metric, gigaflop/s per researcher (unit of measure compute resources), we can see a portrayal of national resources that is not distorted by the scale of the different countries involved. On this basis, in 2009, Canada was 6th in the world. Canada’s international rank (as of 2015) was 24th, out of roughly 34 countries for which data is available. Since 2009, Canada has fallen steadily and is not keeping pace with other G20 countries.
The danger of brain drain is real. Canada will lose both its best research minds and our best expert staff if we don’t provide the resources they want and need. There are plenty of other international organizations eager to attract Canada’s top talent.
Gain infrastructure and operational efficiencies
ARC infrastructure is not something you can buy online. More than 70% of usage today requires capabilities at the leading edge of technology. Compute Canada takes on the work of staying current on fast-changing technologies such as workload portability, storage systems, and data resilience. We also provide technology management coordination between the various Compute Canada sites to ensure that systems are configured and maintained consistently. This ensures that we can provide services to researchers that can run almost anywhere in our platform without hitting infrastructure or programming barriers. We ensure awareness and currency to support global trends towards common services and interoperability.
Compute Canada relieves participating institutions of these headaches. In close cooperation with members, we specify and work with our institutions to procure the systems. We maintain them. We train researchers how to use them. We provide programming assistance and common services. And we have a new Research Data Management and Federated storage service being deployed over the next two years to help researchers store and manage their data and locate and access others’ data.
Together, with our regional partners and institutions, we have a world-class model in place to deliver services nationally and locally. We must ensure our computing capacity is aligned with our research investments. With better coordination, and a funding model reflective of the essential role of advanced research computing, we can achieve and support greater outcomes and accelerated results from our research investments.
Let’s give our smart people the tools they need to compete globally. Let’s invest in world-class advanced computing resources required to support the science we fund. This is innovation’s infrastructure, providing the backbone for 3D manufacturing, curing disease, and understanding our environment, all of which have an immediate impact on our lives, our research community and our economy.
# # #
About Mark Dietrich:
Mark Dietrich is Compute Canada’s President and Chief Executive Officer. He is an entrepreneurial leader with decades of executive management experience, and a proven track record of building and transforming organizations, increasing revenues, and maximizing organizational effectiveness. He specializes in innovative corporate strategy, marketing, business development and exceptional service delivery. Mark was previously the Chief Executive Officer of the Ontario Society of Professional Engineers (OSPE), Ontario’s professional association for engineers. Prior to joining OSPE, Mark founded Bloodstone Solutions Inc., a Toronto-based professional services firm specializing in strategic value innovation, lean services, and eGovernment solutions for public and private sector clients. Mark was also the Executive VP & COO for the Ontario Research and Development Challenge Fund (ORDCF), a $500 million government fund established to support leading-edge, innovative and industrially relevant university and hospital research. The ORDCF provided support to more than 100+ university-industry research partnerships in diverse areas, including nanotechnology, bioinformatics.
Mark’s diverse background in new product development, commercialization, service delivery, and marketing was built through a series of executive leadership roles with companies in the Greater New York City area. His success spans multiple industries, including consulting (strategy, marketing, process improvement, business model innovation), technology and information services (internet service delivery, commercialization, engineered solutions), and finance (project- and asset-based financing), in the United States and Canada, with international mandates. Mark has a BEE degree in Computer Science, Electrical Engineering and Mathematics from Vanderbilt University in Nashville, Tennessee.