Energy Rate Density (φm)
Ergs per second per gram: how intensely a system processes energy
Galaxies
φm approximately 0.5 erg/s/g
Galaxies are vast but diffuse. Despite their enormous total energy output, the energy per unit mass is remarkably low. Most galactic matter sits in dark, cold voids between stars, processing almost nothing. The galaxy is the baseline: complex enough to exist, but barely ticking over.
Sun (average star)
φm approximately 2 erg/s/g
Our Sun fuses 600 million tons of hydrogen per second, yet gram-for-gram it is only about four times as energetic as a galaxy. Nuclear fusion is powerful but spread across an enormous mass. Stars are the universe's first serious complexity engines, but still simple ones.
Earth's climasphere
φm approximately 75 erg/s/g
The climasphere is Earth's thin, restless shell of atmosphere, ocean, and surface: geothermal heat from below, solar absorption from above, circulation and currents in between. It sustains far more complex energy flows than a star because it sits in a gradient: hot core below, cold space above. Gradients are where complexity lives.
Plants
φm approximately 900 erg/s/g
Photosynthesis is a staggeringly sophisticated energy-processing system. A single leaf runs hundreds of chemical reactions in parallel, converting photons into chemical bonds with quantum-level precision. Gram for gram, plants (~900) process energy about 1,800 times more intensely than a galaxy (~0.5).
Animals
φm approximately 20,000 erg/s/g
Animals need locomotion, thermoregulation, immune systems, nervous systems. Every gram of animal tissue is a small factory maintaining homeostasis against entropy. The jump from plants to animals is roughly a 20-fold increase (~900 to ~20,000): the cost of moving through the world rather than sitting in it.
Human Brain
φm approximately 150,000 erg/s/g
The brain is 2% of body mass but consumes 20% of energy. Each gram of neural tissue (~150,000) processes energy about 300,000 times more intensely than a gram of galaxy (~0.5), and 75,000 times more intensely than a gram of stellar matter (~2). This is the cost of cognition: modeling the world, predicting the future, coordinating billions of neurons in real time. Thought is thermodynamically expensive.
Modern Society
φm approximately 500,000 erg/s/g
A modern city is a dissipative structure of extraordinary intensity: power grids, transport networks, communication systems, supply chains. Society is a superorganism whose energy metabolism dwarfs any individual brain. Coordination at this scale requires trust, which is why governance matters thermodynamically.
Computer Chips
φm approximately 10^7 erg/s/g
A modern GPU processes energy more intensely per gram than any known natural structure. This is why data centers need rivers of coolant. We have built matter that dissipates energy about twenty million times more intensely than the galaxy that spawned it: ~10,000,000 erg/s/g against ~0.5. The trend line points upward, and we are building what comes next.
erg/s/g (log scale; bars measured from an origin of 10−1)
What This Measures
Energy rate density (φm) measures how much energy flows through a gram of matter each second. It's a proxy for complexity: more intricate systems require more energy to maintain their structure.
The progression is striking. From a galaxy (~0.5) to a brain (~150,000), φm rises by a factor of three hundred thousand. Your brain, gram for gram, processes 75,000 times more energy than a star (~2).
This is what the universe has been doing for 13.8 billion years: building systems that process energy ever more intensely. We are not an exception to the pattern. We are its latest expression.
The Direction of Cosmic Evolution
From Big Bang to now, average φm has increased. The universe is not just spreading out; it's building structures that process energy more and more intensely.
Complexity Is Expensive
High-φm systems require constant energy input. Turn off the food supply and the brain dies. Turn off the power grid and society collapses. Complexity must be fed.
What Comes Next?
If the trend continues, we should expect even higher-φm systems to emerge. AI may be one. Whatever inherits the future will likely process energy more intensely than we do.