Heat Walls Limit Quantum Computers

In the paper's words

We introduce the thermodynamic wall: a boundary-condition formulation of the trade-off between macroscopic heat dissipation and quantum coherence time in engineered systems. Just as computational wall models in fluid dynamics encode unresolved near-boundary physics into an enriched boundary treatment, the thermodynamic wall encodes an unresolved thermal environment into an effective boundary object that imposes a flux constraint on any quantum device embedded behind it. We develop the framework in two regimes.

The abstract, verbatim, and its opening sentence is the claim. Every line below is quoted from the text.

Numbers it reports

1. In the cryogenic regime, a wall with $G_{\mathrm{th}}=1.0\times10^{-7}\ \mathrm{W\,K^{-1}}$ across $\Delta T=1.0\times10^{-3}\ \mathrm{K}$ removes $P_{\mathrm{wall}}=1.0\times10^{-10}\ \mathrm{W}$, which via Landauer's principle supports at most $f_{\max}=5.22\times10^{8}$ logical cycles per second and a viability window $N_{\max}=5.22\times10^{8}$ coherent operations for $T_2=1\ \mathrm{s}$; the thermal-occupation floor at $5\ \mathrm{GHz}$ and $20\ \mathrm{mK}$ is $n_{\mathrm{th}}=6.16\times10^{-6}$, giving a thermal error time of $16.2\ \mathrm{s}$.

   from Abstract

2. The paper connects to the QNFO research program on thermodynamic viability [10], on ultrametric relaxation dynamics in topological quantum memory [9], and on Planckian dissipation in strongly correlated systems [11]: the thermodynamic wall is the engineered, mesoscopic counterpart of the Planckian dissipation bound that those works locate in correlated matter.

   from 1. Introduction

3. Zhu [6] revises the brick-wall method for computing black-hole thermodynamic quantities: a scalar field's contributions to entropy and other thermodynamic quantities are computed by imposing a cutoff wall in Rindler space, with momentum–frequency relations determined numerically to better than $10\%$ accuracy.

   from 2. Background and Related Work

4. Structural Mediation of Planckian Dissipation in Strongly Correlated Electron Systems [11] locates a universal (Planckian) dissipation rate in correlated matter; we discuss in Section 6 how the Planckian rate would act as a lower bound on the wall's microscopic dissipation channel, tightening our engineered bounds.

   from 2. Background and Related Work

5. It is the engineered counterpart of the thermodynamic-viability thesis of [10] and of the ultrametric relaxation hierarchy of [9], in which slow relaxation channels dominate the long-time budget exactly as $T_{\mathrm{th}}$ dominates here when $n_{\mathrm{th}}$ is exponentially small.

   from 3.5 The viability window

6. Then $$G_{\mathrm{th}} = \frac{kA}{d} = \frac{(1.0\times 10^{-4})\,(1.0\times 10^{-6})}{1.0\times 10^{-3}} = \frac{1.0\times 10^{-10}}{1.0\times 10^{-3}} = 1.0\times 10^{-7}\ \mathrm{W\,K^{-1}},$$ $$P_{\mathrm{wall}} = G_{\mathrm{th}}\,\Delta T = (1.0\times 10^{-7})\,(1.0\times 10^{-3}) = 1.0\times 10^{-10}\ \mathrm{W}.$$ So the wall removes one hundred picowatts — the entire heat budget for erasure.

   from 4.2 Cryogenic regime: full arithmetic

How it is done

1. We derive both answers in closed form and evaluate them for representative parameter sets, with every arithmetic step shown.

   from 1. Introduction

2. Krais, Sander, Beck and Gassner [3] introduced wall modeling for RANS (Reynolds-averaged Navier–Stokes) simulation within a high-order discontinuous Galerkin method: instead of imposing wall functions as boundary conditions, they build the near-wall velocity profile into the function space as a local enrichment, so the Galerkin projection automatically selects the best combination of polynomial and enriched shape functions.

   from 2. Background and Related Work

3. Since $T_2 = 1.0\ \mathrm{s} \lt T_{\mathrm{th}} = 16.235\ \mathrm{s}$, the coherence budget binds: $$N_{\max} = f_{\max}\,T_2 = (5.2247\times 10^{8})\,(1.0) = 5.2247\times 10^{8}.$$ Had $T_2$ exceeded $T_{\mathrm{th}}$, the bound would instead be $N_{\max} = f_{\max}\,T_{\mathrm{th}} = (5.2247\times 10^{8})\times(16.235) = 8.482\times 10^{9}$ operations (arithmetic: $5.2247\times 16.235 = 84.824$, times $10^{8}$).

   from 4.2 Cryogenic regime: full arithmetic

4. With $T_2 = 1.0\ \mathrm{s}$, $N_{\max} = 5.22\times 10^{8}$ coherent logical operations; if $T_2$ were extended beyond $T_{\mathrm{th}}$, the bound would rise to $8.48\times 10^{9}$ operations.

   from 5. Results

What would falsify it

1. We situate the construction relative to wall modeling via function enrichment, the brick-wall model of black-hole thermodynamics, the analytic range of the heat operator, and the QNFO corpus, and we state the conditions under which the resulting bounds would be falsified.

   from Abstract

2. Section 6 discusses limitations and falsification conditions, and Section 7 concludes.

   from 1. Introduction

3. Krank and Wall [7] established the base method for incompressible LES, demonstrating that a problem-tailored function space predicts turbulent boundary-layer gradients on very coarse meshes — the direct ancestor of the "coarse interior, enriched boundary" structure we use.

   from 2. Background and Related Work

Where the claim stops

1. Third, we assumed erasure is the only heat load; coherent control pulses and quasiparticle generation add loads not captured by Eq.

   from 6. Discussion

2. Fourth, the thermal-error model assumes a single thermal bath at $T_c$ coupled at rate $\gamma_c$; nonequilibrium quasiparticle populations can exceed the equilibrium $n_{\mathrm{th}}$ by orders of magnitude, shortening $T_{\mathrm{th}}$ below the computed $16.2\ \mathrm{s}$.

   from 6. Discussion

3. If the wall's microscopic channel were Planckian-limited at this rate, it would be far faster than our flux-limited budget and hence non-binding — but this is a hypothesis imported from [11], not derived here, and it may fail for engineered mesoscopic walls where disorder and isolation suppress the correlated dissipation channel.

   from 6. Discussion

Where to read

1. Abstract205 words · 1 min
2. 1. Introduction461 words · 2 min
3. 2. Background and Related Work824 words · 4 min
4. 3. Methods510 words · 2 min
5. 4. Analysis804 words · 3 min
6. 5. Results292 words · 1 min
7. 6. Discussion480 words · 2 min
8. 7. Conclusion193 words · 1 min