The Thermodynamic Tori

The hypothesis

Bitcoin’s thermodynamic evolution can be modeled as topological phase transitions between torus types. A torus is defined by two radii: R (major radius, distance from tube center to torus center) and r (minor radius, tube radius). The relationship between R and r determines the torus geometry.

R represents the system’s total thermodynamic capacity: the bounded identity manifold (~1.1579 × 10⁷⁷ keys), the fixed supply envelope (21M BTC), and the total observable state-space the protocol can account for. R is protocol-defined and changes only when consensus rules change.

r represents the localized thermodynamic intensity per inscription: the energy expended to resolve one block (subsidy + fees), the fee pressure, difficulty-scaled work, and memory cost per state transition. r is market-driven and variable.

Three regimes emerge:

• Ring Torus (R > r): The container properly encloses its contents. Thermodynamic headroom exists. All costs internalized.
• Horn Torus (R = r): Critical transition. The surface touches itself. Zero headroom. First externalization.
• Spindle Torus (R < r): The surface self-intersects. Hidden volumes appear. Container cannot enclose what it contains. Second externalization.

We know that pre-SegWit Bitcoin operated as a ring torus with constant R/r ratio. SegWit and Taproot changed the rules of how R and r evolve, but we don’t know if the system actually entered horn or spindle torus territory.

3D visualization

Drag the slider to model 100 blocks. At genesis, the torus is at the origin (R = r = 0). As the network grows, it expands into a ring torus. SegWit and Taproot changed the rules of how R and r evolve, but we don’t know if the system actually entered horn or spindle torus territory. Drag to orbit. Scroll to zoom.

Ring Torus (R > r)

Genesis to pre-SegWit (2009–2017). At genesis, the torus was at the origin point: R = r = 0. The supply was 50 BTC, difficulty was 1, and the network was tiny. As the network expanded, both R (total thermodynamic capacity) and r (inscription intensity) grew, but R grew faster than r, so the R/r ratio increased. The original rules enforced R > r by design. The system had thermodynamic headroom.

All costs are internalized. The writer pays for persistence: 1 raw byte = 1 vbyte. The miner pays for resolution. Node costs are priced into fees. The pricing-persistence alignment holds. Every joule expended within the system is accounted for on the observable surface.

The observable surface faithfully encloses what it contains. There is no hidden interior—every point on the surface maps to a unique interior state. The mapping from observable boundary to enclosed state-space is bijective. The container’s surface uniquely determines the set of possible futures.

Horn Torus (R = r)

Unknown — may or may not have appeared. SegWit changed the rules. The witness discount introduced a pricing distortion: witness bytes count as 1/4 weight. The writer pays for 0.25 MB, but the network stores 1 MB. The 0.75 MB gap begins.

Costs are externalized to node runners. The pricing unit (vbyte) diverges from the persistence unit (raw byte). The observable surface is still intact—all scripts remain visible, the bijection holds—but the pricing-persistence alignment breaks.

The R/r ratio changes, but we don’t know if the system actually reached the horn torus (R = r). The surface may touch itself at some point, creating zero headroom. The system can no longer fully account for the costs it creates, but the costs are borne by node runners (physical layer), not validators (informational layer).

Spindle Torus (R < r)

Unknown — may or may not have appeared. Taproot changed the rules again. Taproot introduced committed-but-unregistered states: hidden script trees (MAST), x-only key reflection (2:1 mapping), and silent erasure of prepared possibility. The observable capacity (R) decreased because the state-space is now partially opaque.

Costs are externalized to validators. The true thermodynamic cost exceeds what the on-chain surface reveals. The writer pays for 0.25 MB, the network stores 1 MB, and hidden preparation (MAST tree construction) is real but invisible. The container can no longer enclose what it contains.

R < r becomes possible. The surface self-intersects. Hidden volumes appear—interior regions that are topologically distinct from the exterior but not visible from it. The mapping from observable boundary to enclosed state-space becomes many-to-one. The container’s surface is identical whether it holds one path or a hundred. An epistemic horizon exists within the committed state.

But we don’t know if the system actually entered spindle torus territory. We only know that Taproot changed the rules in a way that makes R < r possible.

Comparison

Note: The visualization above is a model showing how R and r evolve under different rule sets. It does not claim to represent the actual historical trajectory or confirm that the system entered horn or spindle torus territory.

Calculating thermodynamic volume

The torus model gives us a geometric intuition for Bitcoin’s thermodynamic evolution, but it raises a harder question: can we actually calculate the thermodynamic volume of the system? Can we measure R and r in real units and track how they evolve over time?

What we know:

• Identity manifold cardinality (N_priv): ~1.1579 × 10⁷⁷ keys. This is the protocol-defined invariant capacity. It does not change with use or market conditions.
• Block memory surface (τ_t): The physical bytes committed per block. This is observable from chain data.
• Value inscribed (W_t): The satoshis minted or rearranged per block. This is observable from chain data.
• Difficulty-scaled entropy field: The nonce space miners must traverse. This is observable from chain data.

What we might compute:

• Potential thermodynamic volume (V_potential): The total state-space the protocol can account for. This might be N_priv × τ_max × 1 block, where τ_max is the maximum block size under the current rules.
• Realized thermodynamic volume (V_realized): The actual state-space instantiated by the chain. This might be the sum of all UTXO containers: Σ(UTXO_i), where each UTXO is anchored to an identity coordinate, occupies a memory allocation, and contains satoshi capacity.
• Closure ratio (V_realized / V_potential): How much of the potential volume has been realized. This might correspond to the R/r ratio in the torus model.

But there’s a problem. After SegWit and Taproot, we can’t fully compute realized volume from chain data. Hidden states exist: witness data that nodes store but writers don’t fully pay for, MAST trees that are committed but unregistered, x-only keys that fold two identity quanta onto one observable coordinate. The true thermodynamic state exceeds what the on-chain surface reveals.

This is where the model breaks down, or where it becomes most interesting. We have the geometric intuition, but we don’t have the measurement framework. We don’t know how to calculate R and r in real units. We don’t know how to account for hidden states. We don’t know if the torus model is the right way to think about this, or if there’s a better approach.

We’re asking for feedback. Specifically:

• Is N_priv the right measure of thermodynamic volume, or should we use a different quantity?
• How do we account for hidden states (witness data, MAST trees, x-only keys) in the volume calculation?
• Can we compute realized volume from chain data, or do we need external information?
• What’s the right metric to track over time? The closure ratio? The R/r ratio? Something else?
• Is the torus model reasonable, or is there a better way to model Bitcoin’s thermodynamic evolution?

We don’t have answers to these questions. We have a model and a set of open problems. If you have ideas, critiques, or alternative frameworks, we want to hear them. This is work in progress, and we’re building in the open.