More Energy Efficient Computing  

The problem: AI-driven computing is demanding an ever-increasing amount of power. This rising consumption strains power grids and limits the operational life of battery-powered sensors in critical infrastructure.

The solution: National Science Foundation (NSF)-funded research at Carnegie Mellon University focused on fundamentally re-engineering the architecture of a general-purpose computer to prioritize energy efficiency. This effort led to the creation of the most energy-efficient general-purpose processor that has ever existed.

The impact: This technology has been commercialized by the CMU spinout, Efficient Computer, which recently launched its energy-efficient E1 processor and its software stack. Efficient is scaling up its workforce in Pittsburgh and San Jose. Efficient’s architecture, at the heart of E1, reduces energy consumption by 10-100x, reducing operating costs, extending battery life from weeks to years, and virtually eliminating power constraints. Efficient’s technology is poised to enable energy efficiency cost-savings in crucial applications: critical infrastructure, including AI infrastructure, industrial automation systems, robots, automotive systems, space and defense systems, consumer wearables, and many other use cases.

By commercializing foundational research, this innovation provides a path to a future for computing unconstrained by the limitations of energy.

Go deeper: Efficient Computer Saves Energy