By: Dr. Barbara Collignon, Senior Systems Analyst, Bank of Canada.
The views expressed here are those solely of the author. No responsibility for them should be attributed to the Bank of Canada.
On April the 13th 2017, I attended the Data Effect in Ottawa to learn about the next-gen supercomputing Canadian infrastructure. Investments are indeed pouring into AI and HPC, and we can clearly see that “things are moving faster in Canada,” to cite one of the panel chairs, Humza Teherany, CEO of Compass Digital Labs and President, CIO Association of Canada.
The first panel on National Strategy chaired by CEO of Compute Canada, Mark Dietrich, clearly emphasized Canada’s plan to leverage farming through precision agriculture. Already, “Tractors are dragging themselves to plant seed down to millimeters precision,” said Malcom Campbell from the University of Guelph. Farmers and scientists are also collecting real-time data to track a million cows’ daily life producing hundreds of petabytes of data. 500 PB are to be converted into usable resources for the dairy industry and “moving those data from the farms to the computing platform will be the biggest challenge,” said Campbell. John Weigelt from Microsoft mentioned that they are working with partners on White Space WiFi to connect “more centres.” This new shift to bridge farms and researchers is driving a strong need for novel multi-disciplinary teams with broadly-skilled individuals. Campbell illustrated the idea citing his grandpa, who used to be “a farmer, crops specialist, veterinarian and a financier” at the same time.
How do we scale up? “We need people, places and shameless promotion”; “we must avoid start from scratch” and “we must have easier access to Canadian computing resources,” various panelists said. Campbell also mentioned the Canadian Knowledge Mobilization Initiative, which helps bring new tech to the farms. Additional questions were raised at my table. Do (all) farmers really want to share their data? What are the real benefits? How do we secure networks? However, the discussion was not technical, and, in fact, it would have been interesting to discuss next-gen ultra-fast network. Nothing was said either about the need for re-architecting software.
Through the day, the other message was that Canada must go through a cultural change to allow its societal, digital transformation and “stop compartmentalizing people”; “stop being judgmental”; “talk about what you/the government do” and give up on “protectionist attitudes,” especially in academia. Panelist Eyse Bener from Ryerson University talked about her own experience, early in her career, in 1986, as she made some predictions about AI. She was mocked for claiming that, one day, doctors and lawyers would disappear. She now envisions that people will lose their bank account within the next 5 years with the emergence of digital currency. Similarly, panelist Cliff van der Linden, founder of Vox Pop Labs, has been experiencing more resistance in Canada than in Europe or US. His startup uses unconventional methods to help people make decisions that reflect their preferences as “humans are irrational beings,” Cliff said, and do to not always make the best decisions for themselves.
In other terms, “Canada must take risk” to become an “emerging market.” Matthew Mendelsohn from the Government of Canada insisted that Canada must focus on “results and outcomes” instead of “process and rules.”
Deb Matthews, Deputy Premier for the Province of Ontario, also explained that there is a fear that if you release data, people will find out things that you do not want them to find out. She said the government must release data to empower people who work in public services and have their back. She talked about one of the steps she took to help develop a database for storing information about patients, dispensers and prescribers in order to monitor and detect drug abuse. The goal was to help save lives.
Panelist Josée Touchette, COO at National Energy Board, insisted that leadership is important to allow people to create meaning with data. She acknowledged that, “we see a change in the national conversion,” as data made available for visualization are triggering engagement. “Imagination, leadership, hard work and great things happen,” she repeated.
Since the Data Effect conferences, it has been announced that “Canada is quietly adding 10 Petaflops to its network of academic supercomputers” [ref] and Simon Fraser University is now emerging as a major HPC Center. This new shift in canadian HPC leadership suggests that the societal, digital transformation in Canada already started…
Ref : https://www.top500.org/news/canada-is-quietly-adding-10-petaflops-to-its-network-of-academic-supercomputers/
About the Author:
Dr. Barbara Collignon is a French computational scientist who specialized in HPC and mobile development. She was a pioneer in massively parallel molecular docking and her legacy helped significantly advance computer-aided drug design. From 2011 to 2014, she developed and published iPhone apps to promote HPC education and mobile edge computing (initially called “mobile supercomputing”). She currently lives in Canada where she has been helping HPC users implement parallel procedures. In 2017, she joined Bank of Canada as a Senior Systems Analyst to pursue her passion for HPC systems analysis and design. The views expressed here are those solely of the Author. No responsibility for them should be attributed to the Bank of Canada.