Warwickshire’s Conflow Power Group just struck a deal with Nigeria’s Katsina State. Fifty thousand solar-powered lampposts. Each one packing a Nvidia chip. Networked into Africa’s first distributed AI data center. No grid power needed. That’s the pitch. And it’s already moving forward.
Traditional data centers guzzle 300 megawatts from the grid. They demand millions of liters of cooling water daily. Construction drags on for years. Conflow’s iLamps flip that script. A cylindrical solar panel tops each post. It charges batteries. Those feed a 15-watt Nvidia chip inside. Link 50,000 together, and you get 13.75 petaOPS of compute power. Operational from day one. Sun-powered. Grid-free. As Digital Trends reports, this setup sidesteps the massive infrastructure headaches plaguing AI buildouts elsewhere.
But these aren’t just compute nodes. Cameras embedded in the iLamps spot speeding cars. They catch parking violations. Flag seatbelt lapses. Number plate recognition runs in real time. Facial recognition sits on the roadmap—for finding missing people or suspects. Public WiFi and Bluetooth beam out too. In Katsina, traffic fines will fill state coffers. Conflow takes 20% after three years. Rental fees from AI firms using the processing power fund a green bond. That covers installs and upkeep.
Edward Fitzpatrick, Conflow’s chairman, credits Nvidia directly. “NVIDIA is the company that’s created a small enough chip, powered with 15 watts of power, so it can be powered by solar, and we can put that inside a street light,” he told BBC News. Security? Tamper with the chip, and it fries. Posts already run in a UK hospital car park, handling CCTV and plates. Now Katsina gets an assembly factory. Units ship from Morocco, Taiwan, Latvia too.
Dr. Hafiz Ibrahim Ahmad, Katsina’s special adviser on power and energy, calls it groundbreaking. “Home to the only distributed AI data centre of its kind anywhere on the African continent… could mean safer streets, real-time crime and terrorism prevention, free public internet and a revenue stream that flows back into the state,” he said in the BBC piece. Negotiations span seven Nigerian states, universities, institutions. Scale to 300,000 units. Africa’s biggest distributed AI network.
So why Nigeria? Sunshine abounds. Rules bend easier. Fitzpatrick again: “Africa is our prime target because there’s plenty of sunshine which is great, they’ve got more relaxed rules and regulations, they want us to put the street lights on the street.” Florida talks bubble too—with schools eyeing surveillance and interactive features like gesture voting.
Experts temper the hype. Prof. Ian Bitterlin, a data center veteran, flags physical security risks on streets. Communication lags between distant posts kill heavy AI training—like for large language models. John Booth of Carbon3IT agrees. iLamps suit light tasks. Think edge computing access points. Like phone masts feeding bigger centers. They supplement. Don’t supplant.
This lands amid AI infrastructure chaos. Half of U.S. data centers planned for 2026 face delays or cancellation, per a Yahoo Tech report citing Bloomberg. Power shortages. Supply chains snag. Elsewhere, hyperscalers chase wild fixes: SpaceX eyes orbital data centers. Microsoft tested underwater ones. Meta beams space solar. iLamps? Grounded. Practical. Distributed.
Privacy shadows loom. Facial recognition invites bias, misuse. Conflow pledges legal compliance. But streets become eyes everywhere. E-waste warnings grow too—AI strains resources, as Digital Trends notes. Solar changes that math. No rare earths in chips alone. Batteries cycle. Posts endure.
Scale works here. Katsina’s deal proves viability. Revenue sustains it. Edge AI thrives on low latency—lampposts sit where data generates: roads, parks, crowds. Not remote warehouses. Global grids buckle under AI thirst. U.S. operators predict gigawatt shortfalls. Zoning fights erupt, from Wisconsin to Boston. Nigeria sidesteps. Builds on sunlight.
Conflow’s CEO Edward Fitzpatrick frames the shift. “This agreement is a defining moment for how the world thinks about AI infrastructure,” he said in statements covered by Punch Nigeria. Katsina’s 13.75 petaOPS arrives via posts. Sun-fueled. Instant. No 300-megawatt drain.
Critics doubt full replacement. Fair. But for inference? Local analytics? Surveillance feeds? Perfect fit. Multiply by thousands. You cluster compute where needed. Bandwidth bottlenecks ease—process nearby. Global south leads. Others watch. Or catch up.
from WebProNews https://ift.tt/vZIBLcQ
No comments:
Post a Comment