Niagara College viticulture researchers gathering data to help farmers manage their vineyards will be able to more efficiently process larger amounts of information thanks to a donation of high-performance compute power.
The college’s PrAgMatic System, designed to collect field data via sensors and turn it into digestible, usable information that’s specific – whether for a vineyard or vegetable farm – requires a fair degree of processing power, said Mike Duncan, chair of visualization sciences with Niagara College.
“[The processing power] would allow a far more fine-grained computational architecture than, say, a Beowulf cluster would,” he said. And the computational power is necessary considering the unique requirements of individual farmers. “If you have a farmer who is growing tomatoes, their concerns are vastly different from somebody who’s growing beets, and very different from someone growing vines and grapes,” said Duncan.
A cluster of more than a hundred individual computers that originally powered the college’s PrAgMatic System, will be replaced by an nVidia Tesla high-performance GPU processor from Oakville, Ont.-based hardware vendor MDG Computers Canada Inc. Although the hardware is in place, the research team still needs to complete the code development.
The donation was a great way to show interest in a college initiative and “seemed to be a good fit” to showcase the new compute technology that MDG was taking to market, said John Mangold, vice-president for the corporate education division with MDG. The vendor will provide the hardware for the project moving forward.
The research was already well underway before MDG’s donation. The infusion of new technology will simply let researchers continue working as before except with the added capability to process larger amounts of data at a much faster rate, said Marty Jurmain, director of research and innovation with Niagara College.
“Pragmatic will probably never quite be done because technologies that sense better and distribute information to the farmer get smaller and more economical, we’ll take advantages of those,” she said.
The project is also now greener. Replacing the cluster of computers means that the resident 10-ton air-conditioner will no longer need to run at its maximum. “If I had continued along the way I was going, I was going to need another [air-conditioner],” joked Duncan.
As the project progresses, farmers will be able to access the database using an interface that essentially filters data through a set of SQL queries. Currently, only a demo version of the interface exists, but Duncan envisions the interface will present data in either straight text or 3D fly through.
“[Such] analysis may be better presented in 3D than it is in 2D. That will depend entirely on the type of information that is required for that particular farmer,” he said.
But the infusion of new technologies doesn’t just benefit the farmers, the project’s research students get the opportunity to engage in software development they wouldn’t have otherwise had. There are currently a handful of students who work on programming, electronics, and Web design.
And just as the PrAgMatic System is a work in progress, so is the research scope. Right now, it may be focused solely on viticulture, but data will eventually be collected from all farms to benefit all farmers. Duncan envisions colleges across the country collecting local data and producing maps for neighbourhood farms, which are then amalgamated into a national “uber system”.
If the project does take on a national scope, the scalability of the nVidia Tesla processor units will certainly prove handy for expanding operations, said Duncan. “But my focus right at the moment is to get it right on a single farmer’s field, then for two farmers’ fields, then for many farmers’ fields.”
Mangold agreed that, over time, the technology will become widely available to all farmers. But for Niagara College, in particular, situated in the heart of wine country, it’s “a chance for them to put a stake in the ground, and take the technology and research and make farming more productive.”
