Q&A with Nizar Ladak, CEO of the Digital Research Alliance of Canada, on integrating digital research infrastructure
October 5, 2021
In 2019, Innovation, Science, and Economic Development Canada created the New Digital Research Infrastructure Organization (NDRIO), a national not-for-profit with a mandate to create a broad and integrated Canadian digital research infrastructure (DRI) ecosystem. Two years on, NDRIO is on the verge of launching its inaugural funding opportunities and a five-year strategic plan. It has also just announced a new name: the Digital Research Alliance of Canada.
Research Money spoke with the Alliance’s CEO Nizar Ladak, who was formerly the president and CEO of Compute Ontario. We discussed the new organization's journey and how it can ensure that Canadian researchers have the digital tools they need to tackle society’s problems today and in the future.
Research Money: I'd like to set the stage by first asking: why was a new digital research organization needed?
Nizar Ladak: With the proliferation of big data, researchers need "big compute" to be able to process the research questions that are going to change humanity. But the tools they need to place that data upon are extraordinarily expensive. To give you a comparator, the last supercomputer Canada built, at the University of Toronto, cost $26 million, not including the support costs, which are in the multiple millions of dollars. So the only way for Canadian researchers to be able to truly compete at a global scale is through the sharing of costs. And so our organization was formed to oversee the federal coordination and activity that happens for any researcher in a post-secondary institution to access this machinery.
R$: The organization is designed to oversee all three pillars of digital research infrastructure, advanced research computing, research software and research data management, which is a shift in direction for Canada. What are the advantages of bringing these pillars together?
NL: We serve about 8,000 post-secondary researchers in Canada. Many would argue that keeping them separate has nothing to do with the way in which those tools are used. If you think about it intuitively, most researchers say, "I have to have the big machines, I have to have software to put on those machines, and how can I not organize my data according to common frameworks?" So the fact that they emerged as separate entities is an artifact of who was willing to fund them. So we’re adopting what I would call a researcher-centric orientation, bringing together those three areas that were formerly siloed.
R$: You've been doing extensive consultations over the last year or so, including a needs assessment that was completed by more than 1,300 researchers. What did you hear from the research community about their needs?
NL: The key message that researchers delivered is that the need to access these tools to do their work is only going to grow exponentially.
Our analysis also revealed that only 40 per cent of the demand that's been expressed by the 8,000 researchers across the country is being met. So what are the other 60 per cent doing? That's what necessitated the creation of our organization. Your readers will be very disappointed to hear that in the absence of these tools, three things were happening. One, Canadian researchers were partnering with other countries just to run the research and giving the first authorship for their intellectual property. Two, they were paying Amazon, Google and others at about four times what it would cost to get access to clusters at a not-for-profit centre. And three, they were actually at times so frustrated from failing to gain access, they would sell their intellectual property on the open academic market. So that’s why the government stepped in.
R$: Where are the opportunities for transformation in the digital research infrastructure ecosystem that we have today?
NL: I would say, we're not behaving like an ecosystem, like an interdependent set of partners. That's entirely driven not by the will of the people who support the researchers as much as the funding mechanisms that have motivated those behaviours. As a result of those mechanisms, we weren't seeing a proliferation of best practices, we weren't seeing cooperation, we were seeing the bare minimum to fulfill needs with the anticipation that comes around the next competition.
We are aiming to change that. We are going to be creating a process where we will be awarding $60 million of our money, plus $40 million from the provinces and territories, for a total of $100 million annually to rejuvenate this infrastructure. That process will not necessarily be organized through a competition but will instead be informed by need. We'll be asking questions like, of the five data centres, whose equipment is the oldest and which needs to be refreshed? Are there particular clusters that we may need to reserve, given the proliferation of needs that exists among machine learning and AI scientists or astrophysicists or medical professionals or the digital humanities? Who's best-suited to host and support that equipment? So we have visions or aspirations of a common table where we can all come together and say, what's the right decision for Canada versus who competed best and put in the best proposal? Those are some of the cultural changes we hope to institute.
R$: We've talked about the vision and the opportunities for transformation. How do those translate into programs, projects and next steps for the organization?
NL: From an operational point of view, our focus right now is bringing multiple organizations together, specifically Compute Canada which oversees advanced research computing. We have a massive transition ahead and that is our principal focus between now and the end of the fiscal year.
Strategically, the board identified 32 objectives that they hope to achieve over the next five years and close to 200 ideas about how to accomplish those 32 strategic objectives. Now, we're taking those back to consultation and saying, from the needs assessment and the current state assessments, this is what the community told us. Do any of these 200 ideas have the merit to actually become projects that we could actually accomplish and tie to the achievement of these strategic objectives?
Both the results of the needs assessment and the strategic plan will be publicly available on our website shortly.
R$: NDRIO has a new name, the Digital Research Alliance of Canada, or the Alliance. What do you want it to convey to the research community?
NL: What we wanted to convey through our name, and hopefully through our actions and our approaches and behaviours, is that in order for Canada to punch above its weight, we need to come together as provinces and territories and give our researchers the tools to do their work. The Alliance is a like-minded, cooperative collaborative of people who wish to elevate Canada to the global stage of the knowledge economy. That’s what we need because Canada doesn’t have the spending power of the United States and China.
R$: Canada's digital research infrastructure has been deemed “fragmented, oversubscribed and underfunded.” If you could choose three adjectives to describe it in five years, what would they be?
NL: Actually, during my listening tour across Canada last year, I pointed to three nouns that used to define our ecosystem. Those three nouns were segregation, division and competition. The new nouns that I hope we will aspire to are unity, integration and collaboration.