[ ENERGY ]
Second Law of Thermodynamics
The second law tells us that sustaining order and processing information come with energy cost and entropy increase. Civilization and technology are, in essence, local creation and maintenance of order; science is seeking more efficient, sustainable paths in theory and engineering. The following directions combine physical depth and societal meaning.
In theory, reversible computation without information erasure can approach zero energy. Landauer’s principle ties erasure to minimum cost kT·ln(2). Exploring reversible and near-reversible systems advances low-power chips and data-center efficiency and deepens our understanding of information and thermodynamics.
Qubits in superposition carry far more information density than classical bits; the brain achieves complex cognition at ~20 W. Studying energy cost of quantum algorithms and neuromorphic, event-driven architectures points the way to next-generation efficient computing and AI hardware.
Maxwell’s demon and information engines show that information can act as “negative entropy” in the conversion of energy and order. Clarifying the relation among information, energy, and entropy has fundamental value and a unified view for energy management, intelligent systems, and sustainability.
From solar and wind to fusion, from batteries to hydrogen, humanity is broadening energy sources and improving conversion and storage. Each breakthrough directly affects climate security, energy equity, and global development—among the most socially valuable frontiers in engineering and science.
Information erasure has a theoretical minimum cost, but reversible or near-reversible systems are extremely hard to realize in practice. Can we approach this limit for computation and communication while respecting thermodynamics? This concerns both chip and data-center efficiency and whether civilization can sustain itself on finite energy.
Large-scale AI training and massive data processing have drawn attention for energy and carbon footprint. How to increase capability while controlling energy and environmental cost? Algorithmic innovation, hardware efficiency, and clean energy together are a shared challenge for academia and industry.
Solar and wind are intermittent; storage and grid dispatch are critical. Making clean energy accessible and affordable globally involves technology, policy, and equity and requires cross-disciplinary and international collaboration.
Fusion, next-generation storage, and hydrogen keep advancing in the lab but face cost, safety, and infrastructure challenges at scale. How to accelerate maturation and deployment so all of humanity can share clean, stable, affordable energy is a defining task for science and engineering.
Plasma physics, tokamak and stellarator design, first-wall materials, tritium cycle and safety. If controlled fusion is achieved, it could provide nearly unlimited clean baseload energy with profound implications for climate and sustainability.
Solid-state and flow batteries, hydrogen production and storage, compressed air and novel storage. Cost-effective, long-lived, safe storage is key to large-scale renewables and electrified transport, and to emissions reduction and energy security.
High-efficiency PV and solar thermal, offshore wind, geothermal and marine energy; new semiconductors and materials for conversion. Lowering levelized cost and improving reliability so clean power reaches more regions and people.
Multi-energy complementarity, demand response, distributed and microgrids, energy internet. Digitalization and intelligence to improve efficiency and integrate variable renewables while ensuring security and fairness—infrastructure for a low-carbon society.
Overcoming plasma instability, high-temperature materials, tritium self-sufficiency and safety to reach net energy gain and engineering feasibility—one of the most challenging and consequential goals in the history of energy.
Developing low-cost, long-life, safe large-scale storage to address the intermittency of renewables and deliver reliable clean power 24/7 for industry, transport, and households.
From chips and algorithms to data centers, systemically reducing energy and carbon per unit of compute. In the AI and digital era, efficient computing is both a business imperative and essential for global decarbonization and sustainability.
Making clean, affordable energy available to developing countries and vulnerable groups where technically and economically feasible. Science, engineering, and policy together are needed for an energy transition that leaves no one behind.