AI is happening at the edge — in laptops, phones, cars, cellular base stations, and even network devices like routers. Processing AI close to the user can make it less expensive, reduce energy consumption, and unlock new capabilities.
When the cloud infrastructure boom happened, only a select few companies with deep pockets for server farms — Amazon, Microsoft, Google — were able to dominate. But AI's evolution is decoupling power from sheer size. "Doing more with less" has become the mantra, with smaller hardware able to support powerful models, and better AI outcomes not necessarily requiring more raw processing power.
We spoke with Brent Summers, Senior Marketing Manager at Qualcomm Technologies, to discuss how AI is proliferating not just in the cloud but on the edge, across networks, devices, and cars.
A virtuous cycle: "Innovation is a virtuous cycle," Summers explains. "Hardware innovation enables new software capabilities. As software companies develop those capabilities, they drive new hardware requirements. AI is already proliferating across every point in the network."
Edge opportunities: While most attention focuses on AI in the cloud, Summers observes that the technology's reach is far deeper. "AI is happening at the edge — in laptops, phones, cars, cellular base stations, and even network devices like routers," he says. "Processing AI close to the user can make it less expensive, reduce energy consumption, and unlock new capabilities."
3G, 4G, and 5G: Unlike many players in the space, Qualcomm Technologies has purview over both owned-device ecosystems and broader infrastructure support. "We are a fundamental technology player — a hardware company enabling new software experiences," Summers notes. The company's position is reinforced by decades of investment: "$100 billion in R&D and 160,000 granted patents and pending patents." Known for breakthroughs in 3G, 4G, and 5G, Qualcomm Technologies has also built a strong foundation in AI, especially at the edge, where the future of AI is taking hold.
Unit economics: Doing more at the edge, Summers believes, is poised to be the next major frontier for AI. "It actually changes the unit economics for a SaaS company when they can manage workloads on your phone or laptop instead of in their data center, where they're paying for it."
As someone committed to sustainability, I believe in the transformative potential of AI—but must also acknowledge the challenges it poses to a sustainable future. The shift toward on-device AI processing is promising because it enables novel use cases and can be more energy efficient.
Energy, time, resources: Although the cloud still plays a critical role, Summers acknowledges a meaningful portion of AI workloads will soon run directly on devices. "We're already starting to see AI workloads move from the cloud to the edge," he continues. "Copilot+ PCs are a good example of that, where you can actually run the same kinds of AI tasks without connecting to the cloud, and it'll work in airplane mode. There's still absolutely a place for the cloud, but there's a transaction cost associated with sending it to the cloud, hitting the data center, keeping the data center cool, sending the query, and then getting the response back to you. That is energy, time, and resource costs."
Cloud vs. phones: Beyond cost and performance benefits, Summers points to another important reason edge AI matters: energy efficiency and sustainability. "You’ve probably seen stats about generating an image with some AI image generators — it’s like pouring out a bottle of water, or it takes the same charge as a cellphone," he says. "Research shows that doing the same activity — generating an image with Stable Diffusion — on a phone equipped with a Snapdragon processor is 30.60x more efficient on a per kWh basis than data center generation."
ESG goals: The implications of AI use for corporate sustainability efforts — and the planet — are massive. "As someone committed to sustainability, I believe in the transformative potential of AI—but must also acknowledge the challenges it poses to a sustainable future," Summers reflects. "The shift toward on-device AI processing is promising because it enables novel use cases and can be more energy efficient."
