Nicolas Barrette to explore the architectural obsolescence of data centers as 2026-27 Banham Fellow
Kelly Sheldon July 31, 2026
Nicolas Barrette has been selected as the 2026-27 Peter Reyner Banham Fellow at the UB School of Architecture and Planning.
Nicolas Barrette was in his early teens when he picked up The Cuckoo’s Egg, Clifford Stoll’s memoir that recounts the search for the hacker who infiltrated the computer system at the Lawrence Berkeley National Laboratory. The book quickly became his favorite – a “strange choice for a 13-year-old,” he admits – sparking a lasting fascination with technology. What followed were years spent tinkering with computers, building his own machines, and learning about the systems that operate them.
“What fascinated me was the physicality of it,” he explains. “The cables, ethernet cards, motherboards, and computer towers made the internet feel less like an abstract space than a material system with its own architecture."
Yet when Barrette entered college, he chose a different path, first earning his bachelor’s degree in geography – urban systems from McGill University and then his Master of Architecture from the University of Toronto.
Over the next decade, he built a career in architecture, specializing in projects involving adaptive reuse, heritage transformation, climate-responsive design, and the economic and material systems that shape the built environment. Along the way, he managed to smuggle some of his early technological interests into his work, but now, they’ve become a central focus.
Barrette served as Heritage Lead on a project at 197 King Street East, Old Toronto. Photo: Oldstonehenge
That convergence of architecture and technology will take center stage this year as Barrette joins the UB School of Architecture and Planning as the 2026-27 Peter Reyner Banham Fellow. Named for the influential architectural historian who taught at UB from 1976-80, the fellowship honors Banham’s influential scholarship on material and visual culture as a reflection of contemporary social life.
As Banham Fellow, Barrette will conduct research, teach students, and present his findings to the public, all focused on the topic of Data Centers, AI, and the Architecture of Anticipated Vacancy in Buffalo.
Two things in particular drew Barrette to this fellowship: its unique integration of research, teaching, and public engagement, and the legacy of Banham himself.
“He had a remarkable ability to take seriously the things architecture dismissed,” Barrette notes. “For instance, the sheds, the machines, the gizmos, the environmental systems, the ordinary and overlooked architectures that revealed something larger about culture and technology—he approached them with curiosity, genuine affection, and, when warranted, a caustic edge. My research asks architecture to pay attention to a building type it has largely chosen not to see, so pursuing this work under Banham’s name, at an institution where he taught, feels like a natural continuation.”
For Barrette, Buffalo provides the ideal setting for this research. “There is no closer laboratory for this work than Buffalo,” he observes. “There’s something fitting about studying a supposedly borderless cloud from a border city, where the infrastructure of two countries meets.”
The city’s grain elevators, once symbols of industrial innovation, now largely stand empty as monumental remnants along the Buffalo River. They offer a compelling parallel to today’s data centers. Like those silos, data centers are being constructed at a massive scale despite the likelihood of rapid obsolescence. The technology they house evolves so quickly that facility requirements can become outdated even before construction is complete. The result is a building that’s been designed with little to no involvement by architects and no plan for future reuse.
"This is the friction between the fixity of construction in the material world and the velocity of compute in the platform world," Barrette says. By some accounts, there is roughly a trillion dollars of data center infrastructure in the United States, and more capacity has been built in the past three years than the Eisenhower interstate system managed in forty, with trillions more planned.”
Banham had a term for this condition, acute obsolescence, and it is why Barrette finds his thinking urgent now.
“This fellowship is the first opportunity where the threads fully tie together as it relates to computing, architecture, infrastructure, and the afterlife of technological systems,” Barrette reflects. Out of that consolidation came a central question: What happens to the infrastructures being built today once they too, inevitably, become artifacts?
“We were taught to see cloud computing as weightless, placeless, immaterial, nothing like the heft of grain silos and factories. It has turned out to be one of the largest building programs of the 21st century, manifest as a planetary archipelago of windowless sheds consuming land, water, and electricity at industrial scale, designed almost entirely without substantive input from architects.”
From Barrette's thesis research: Shanghai, China. Photo: Nicolas Barrette
Barrette plans for his work to culminate in a publication that reframes the cloud as a territorial and design problem. The project has three main goals:
1. Establish an architectural history of the data center in North America, situating it in a lineage of machine architecture that also runs through Buffalo itself, from the grain elevator to the Larkin complex to contemporary server farms.
2. Apply architectural representational and mapping tools, beginning with a territorial analysis, tracing a single building’s connections to land, energy, water, labor, communities, and construction.
3. Address the question of afterlives. If these structures are obsolescent by design, then decommissioning, material recovery, electronic waste, and the brownfields they leave behind must be thought through now. Adaptive reuse belongs in that conversation from the start.
Students will play a major role in that work through a studio focused on the afterlife of a hypothetical data center. They’ll be tasked with imagining the possibilities for these buildings—as ruins, archives, mines, or spaces capable of hosting entirely new programs. To ground those proposals, they’ll first study earlier generations of these types of structures, like grain elevators, mills, power stations, and big-box retail stores, while also developing an understanding of how modern data centers operate.
“The goal is to ground work in the ways large-scale infrastructures have been inherited, adapted, reused, or abandoned,” Barrette explains. “I’ve learned that students do their best work when the problem is real and they have genuine agency—essentially, when their work is not simply an exercise but a contribution in its own right.”
He also believes that today’s students bring a perspective that previous generations cannot. Having grown up immersed in digital environments, they possess unique insights into how these technologies shape culture, society, and the built environment.
Barrette served as Heritage Lead on a project at the former IBM Headquarters in Toronto, including selective demolition. Photo: Nicholas Barrette
Rather than guiding students toward a predetermined conclusion, Barrette hopes to encourage critical engagement. “I want them to at least recognize that compute infrastructure is ultimately a territorial question, and that even if the last data center were built tomorrow, architects could still have an essential role in shaping what these buildings, and their afterlives, become.”
His teaching philosophy is informed by previous experience at the University of Toronto, where he learned to use archival research, mapping, photography, and writing as design tools. These approaches help his students understand how past decisions continue to shape present conditions and reveal the layers of material, cultural, and environmental intelligence embedded in existing landscapes and buildings — “precisely the affordances offered by data centers,” he points out.
The timing of the fellowship feels especially significant. Having recently concluded a decade-long chapter with a heritage architecture practice, Barrette sees this as an opportunity to reconnect with his longstanding interest in computing at the very moment when it’s being reshaped dramatically by artificial intelligence.
“Disciplinarily, the alignment is hard to overstate,” he shares. “A century ago, Buffalo's grain elevators were the data centers of their day, functioning as automated, windowless bulk-storage machines sited around cheap power and a logistics corridor. They were admired by European modernists, largely overlooked by American architects, and then rendered obsolete almost overnight. Banham, teaching at UB, read those elevators as A Concrete Atlantis. I wonder if today's cloud is a Silicon Atlantis in the making.”
As Banham Fellow, Barrette returns to the physical nature of computation that grabbed his attention decades ago. Computing has always had a physical footprint, evolving from machines hidden university basements to vast data centers that occupy buildings the size of city blocks and beyond. Yet despite their scale, these systems often remain out of sight and out of mind, a perception shaped by the tech industry and largely embraced by architecture itself.
“One of the most consequential building programs of the century has emerged as expendable architecture built at a permanent, landscape-altering scale. Few have asked what it will leave behind."
